Power supply quality full-function regulation device and multi-time scale optimization control method thereof

Abstract

The application discloses a power supply quality full-function regulation device and a multi-time scale optimization control method thereof. In a normal grid-connected operation mode, a double-layer collaborative control based on long-short time scale decoupling is executed, global optimization scheduling is carried out based on a capacity reservation mechanism in a long time scale, a network side cross-port load sharing system overload is constructed and transient support is provided in a short time scale, and power quality comprehensive compensation is executed on-site with a network side; in a fault off-grid operation mode, time sequence collaborative control oriented to system survival period is converted, a network side is transitioned into a dominant voltage source to establish an island benchmark, a network side keeps continuous tracking on bus voltage, hierarchical load shedding protection is executed after an island enters a steady state, seamless grid-connected reset is executed after an upstream power grid is restored and meets a synchronous parallel condition, and global optimization scheduling is re-accessed. The application can realize comprehensive management of power quality of a distribution area and high-reliable power supply in all working conditions.

Description

Technical Field

[0001] This invention relates to the field of power quality problem management technology, and in particular to a full-function power quality control device and its multi-time-scale optimization control method. Background Technology

[0002] In recent years, with the increasing proportion of new energy power generation connected to the distribution network, coupled with the connection of nonlinear and impulsive loads in power electronics, the distribution network has been transformed into an active network, resulting in more complex power flow distribution and prominent power quality problems such as short-term heavy overload of distribution areas, three-phase imbalance, voltage fluctuations, and harmonics. On the other hand, as the national economy's dependence on electricity has increased year by year, voltage dips caused by distribution network faults or industrial load fluctuations have seriously affected the normal operation of important loads, causing huge economic losses.

[0003] Against this backdrop, power quality management equipment technologies, including Dynamic Voltage Restorer (DVR), Uninterruptible Power Supply (UPS), and Static Var Generator (SVG), have matured and are widely used in various fields. However, due to limitations such as operation and maintenance costs and limited functionality, current technical solutions cannot achieve full-function management of complex power quality problems, nor can they guarantee high-quality power supply under fault conditions. Summary of the Invention

[0004] This invention provides a power quality full-function control device and its multi-time-scale optimization control method to solve the defects of existing power quality management equipment, such as single function, inability to achieve full-function management of complex power quality problems, and inability to take into account high-quality power supply under fault conditions. It realizes comprehensive management of power quality in distribution substations and high-reliability power supply under all operating conditions.

[0005] This invention provides a multi-timescale optimization control method for a full-function power quality control device. The full-function power quality control device includes: a grid-connected converter, a grid-connected converter, a hybrid energy storage system, a three-winding transformer, and a controller. The grid-connected converter and the grid-connected converter share a DC bus, and the hybrid energy storage system is connected to the DC bus. The AC side of the grid-connected converter is connected to the load winding side of the three-winding transformer, and the AC side of the grid-connected converter is connected to the compensation winding side of the three-winding transformer. The controller is signal-coupled with the grid-connected converter, the grid-connected converter, and the hybrid energy storage system. The multi-timescale optimization control method is executed by the controller. The multi-timescale optimization control method includes: a normal grid-connected operation mode and a fault-off-grid operation mode.

[0006] In the normal grid-connected operation mode, a two-layer collaborative control based on decoupling of long and short time scales is implemented: On the long time scale, global optimization scheduling aimed at distribution area energy management and peak shaving / valley filling is performed, including: solving for a steady-state reference operation command set based on a pre-built global optimization scheduling model containing a dynamic capacity reservation mechanism; wherein, the dynamic capacity reservation mechanism is used to limit the upper limit of output of the grid-type converter and the grid-following converter in positive-sequence fundamental wave energy scheduling, so as to forcibly reserve dynamic capacity margin for transient inertia response and power quality compensation; on the short time scale, the grid-type converter is controlled based on a virtual synchronous generator mechanism to adjust its output according to different operating conditions to support the voltage quality of the distribution area; while controlling the grid-following converter to execute the steady-state reference operation command set, the current of the grid-following converter is adjusted to perform comprehensive power quality compensation locally;

[0007] In the off-grid fault operation mode, time-series coordinated control guided by ensuring system lifetime is executed: when a grid anomaly is detected, the grid-connected circuit breaker is disconnected, wherein the global optimized scheduling is correspondingly suspended, and the distribution substation forms an islanded microgrid; during the transient phase of the islanded microgrid, the grid-type converter is controlled to smoothly switch to the dominant voltage source, and the amplitude, phase, and frequency reference of the islanded microgrid bus voltage is established; the grid-following converter is controlled to lock the phase and frequency of the islanded microgrid bus voltage in real time through a phase-locked loop, while simultaneously acquiring and tracking its amplitude; after the islanded microgrid enters a steady state, based on the pre-constructed islanded coordinated optimized scheduling model and the real-time status of the hybrid energy storage system, graded load shedding protection is executed; when the upstream grid recovers to a preset qualified state and meets the synchronous parallel connection conditions, the grid-type converter and the grid-following converter are controlled to perform grid reset and reconnect to the global optimized scheduling.

[0008] In some embodiments, constructing a global optimization scheduling model including a dynamic capacity reservation mechanism specifically includes: establishing a comprehensive objective function, wherein the comprehensive objective function takes minimizing steady-state voltage deviation, optimizing the power factor at the point of common coupling, reducing system active power losses, and optimizing the state of charge of the hybrid energy storage system as joint optimization objectives; establishing system steady-state operation constraints, wherein the system steady-state operation constraints include: setting the upper limit coefficient of steady-state capacity utilization of the grid-type converter and the upper limit coefficient of symmetrical fundamental frequency capacity utilization of the grid-connected converter respectively; and constructing the global optimization scheduling model based on the comprehensive objective function and the system steady-state operation constraints.

[0009] In some embodiments, controlling the grid-connected converter to adjust its output based on a virtual synchronous generator mechanism according to different operating conditions on a short timescale to support the voltage quality of the distribution substation specifically includes: when the distribution substation is under heavy overload and the steady-state output of the grid-connected converter is limited, based on the virtual synchronous generator mechanism and combined with the real-time capacity status of the grid side, locally correcting the steady-state reference operating command set, controlling the grid-connected converter to inject steady-state active current across ports to compensate the three-winding transformer. The load winding side voltage drops; under the condition that the distribution substation is not under heavy load and the upstream grid voltage is distorted, based on the virtual synchronous generator mechanism, the negative sequence and harmonic distortion components of the upstream grid voltage are extracted, and the grid-type converter is controlled to superimpose the corresponding compensation voltage to perform bus voltage waveform repair; under the condition that the upstream grid experiences transient disturbances, based on the virtual synchronous generator mechanism, the dynamic capacity margin is invoked, and the grid-type converter is controlled to autonomously output transient power compensation in response to the frequency and voltage changes of the upstream grid.

[0010] In some embodiments, while controlling the grid-connected converter to execute the steady-state reference operation command set on a short time scale, adjusting the current of the grid-connected converter to perform comprehensive power quality compensation locally includes: extracting the negative sequence current component, harmonic current component, and local high-frequency power deviation component from the load winding side of the three-winding transformer, and performing vector synthesis with the steady-state reference operation command set to generate a comprehensive reference command; applying a phase-segment current safety clamping mechanism to the comprehensive reference command, prioritizing the closed-loop tracking of the steady-state reference active power command when the single-phase current approaches the physical safety limit, and dynamically reducing the amplitude of the power quality compensation command, so as to adjust the output current of the grid-connected converter in a closed loop to perform comprehensive power quality compensation locally, wherein the comprehensive power quality compensation includes: performing high-frequency smoothing of three-phase imbalance, waveform distortion, and local transient fluctuations.

[0011] In some embodiments, the method further includes: adjusting the virtual moment of inertia and damping coefficient of the grid-type converter based on the angular frequency deviation and rate of change of the islanded microgrid, specifically including: extracting the angular frequency deviation and rate of change of the islanded microgrid in real time; dynamically adjusting the virtual moment of inertia of the grid-type converter according to the product of the angular frequency deviation and the rate of change of the angular frequency: increasing the virtual moment of inertia of the grid-type converter when the product is greater than zero; reducing the dynamic virtual moment of inertia to a preset steady-state baseline value when the product is less than or equal to zero; and dynamically adjusting the damping coefficient of the grid-type converter based on the absolute value of the angular frequency deviation and the absolute value of the rate of change of the angular frequency.

[0012] In some embodiments, constructing an islanded collaborative optimization scheduling model specifically includes: establishing an objective function for the islanded microgrid, wherein the objective function takes minimizing steady-state voltage deviation, minimizing tiered load shedding penalties, and optimizing the state of charge of the hybrid energy storage system as joint optimization objectives, and the tiered load shedding penalties introduce node load importance penalty coefficients and shedding ratio optimization variables; reconstructing the constraints of the islanded microgrid: configuring the grid-connected converter and distributed generation as controlled power nodes to bear the islanded basic load, configuring the grid-connected converter as an islanded balancing node to make up for the power difference, and configuring the hybrid energy storage system to maintain the voltage stability of the DC bus; and constructing the islanded collaborative optimization scheduling model based on the objective function and the constraints.

[0013] In some embodiments, after the islanded microgrid enters a steady state, tiered load shedding protection is executed based on a pre-built islanded collaborative optimization scheduling model and the real-time status of the hybrid energy storage system. Specifically, this includes: periodically calculating the expected maximum power support time of the islanded microgrid based on the real-time state of charge, safe state of charge limit, rated total energy storage capacity, and current filtered steady-state discharge active power of the energy storage units in the hybrid energy storage system; and updating the islanded load importance priority list according to the solution results of the islanded collaborative optimization scheduling model. When the expected maximum power support time is lower than a preset time safety threshold, or the real-time state of charge is lower than a preset power warning threshold, tiered load shedding protection is triggered, and non-critical load nodes are disconnected in the order of the islanded load importance priority list.

[0014] In some embodiments, when the upstream power grid recovers to a preset qualified state and meets the synchronization and paralleling conditions, the grid-type converter and the follow-grid converter are controlled to perform grid-connection reset and reconnect to the global optimized scheduling. Specifically, this includes: before closing the grid-connected circuit breaker, activating the pre-synchronization compensation control loop of the grid-type converter, using the phase and amplitude compensation commands between the upstream power grid and the islanded microgrid as feedforward fine-tuning signals, and superimposing them into the virtual rotor motion equation and excitation control equation to eliminate deviations; after meeting the synchronization and paralleling conditions and closing the grid-connected circuit breaker, the grid-type converter is controlled to reset the steady-state reference command set and redetermine the dynamic capacity margin, and the follow-grid converter is controlled to recover to the controlled current source mode and reconnect to the global optimized scheduling.

[0015] This invention also provides a full-function power quality control device, comprising: a grid-connected converter, a grid-connected converter, a hybrid energy storage system, a three-winding transformer, and a controller; the grid-connected converter and the grid-connected converter share a DC bus, and the hybrid energy storage system is connected to the DC bus; the AC side of the grid-connected converter is connected to the load winding side of the three-winding transformer, and the AC side of the grid-connected converter is connected to the compensation winding side of the three-winding transformer; the controller is signal-coupled with the grid-connected converter, the grid-connected converter, and the hybrid energy storage system; the controller is configured to execute the multi-time-scale optimization control method described above.

[0016] In some embodiments, the power quality full-function control device adopts a back-to-back power electronic converter topology with a common DC bus. The AC side of the grid-connected converter is connected to the load-side winding of the three-winding transformer via an inductance-capacitance-inductance (LCL) filter, and the AC side of the grid-connected converter is connected to the compensation-side winding of the three-winding transformer via an inductance-capacitance (LC) filter. The hybrid energy storage system includes a power-type energy storage unit and an energy-type energy storage unit, which are connected to the DC bus via a direct current-to-direct current (DC / DC) converter.

[0017] The power quality full-function control device and its multi-time-scale optimization control method provided by this invention have the following technical effects: 1. A two-layer collaborative control architecture with decoupled long and short time scales is constructed. The upper layer uses global optimization scheduling on a long time scale to formulate a baseline operating strategy to achieve efficient energy management and peak shaving in the distribution area. Combined with real-time disturbance perception and dynamic compensation on a short time scale at the lower layer, transient support and local power quality management are achieved. This architecture breaks through the limitations of traditional single-point management and single-time-dimensional control of converters, realizing the physical separation and logical unification of macro-energy scheduling and micro-power quality management. It significantly improves the operational stability of active distribution networks under varying operating conditions and substantially reshapes the impedance characteristics of distribution nodes, enhancing the support rigidity of local weak power grids. 2. Full-condition system-level multi-objective collaborative control of grid-connected converters, grid-linked converters, and their downstream loads is achieved. Under normal grid-connected conditions, the potential of the grid-connected and grid-connected converters is fully utilized, with no idle converters. Through the coordinated operation of the grid-connected and grid-connected converters, high-quality power supply to the load is achieved. Under grid fault off-grid conditions, the grid-connected converters establish an island voltage reference, and the grid-connected converters maintain continuous tracking of the bus voltage, ensuring the transient stability of the islanded microgrid and the highly reliable and continuous power supply to critical and sensitive loads. Attached Figure Description

[0018] To more clearly illustrate the technical solutions in this 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 some embodiments of this invention. For those skilled in the art, other drawings can be obtained from these drawings without creative effort.

[0019] Figure 1 This is a schematic diagram of the multi-time-scale control hierarchy of the power distribution network for the full-function power quality control device provided by the present invention.

[0020] Figure 2 This is a schematic diagram of the back-to-back converter main circuit topology of the power quality full-function control device provided by the present invention.

[0021] Figure 3 This is a schematic block diagram of the hybrid energy storage adaptive power allocation provided by the present invention;

[0022] Figure 4 This is a schematic block diagram of the virtual synchronous generator of the grid-type converter provided by the present invention;

[0023] Figure 5 This is a schematic block diagram of the grid-connected converter provided by the present invention;

[0024] Figure 6 This is a flowchart illustrating the multi-timescale optimization control method of the power quality full-function regulation device provided by the present invention.

[0025] Figure label:

[0026] 100: Full-function power quality control device 110: Grid-type converter 111: LC type filter 111A: Second filter capacitor 111B: Filter Inductor 120: Grid-type converter 121: LCL type filter 121A: First filter capacitor 121B: Inverter-side inductor 121C: Grid-side inductor 130: Hybrid Energy Storage System 131: Energy Storage Unit 132: Power-type energy storage unit 133: DC / DC converter 133A: Supercapacitor DC / DC Controller 133B: Battery DC / DC Controller 140: DC bus 141: First capacitor 142: Second capacitor 150: Three-winding transformer 151: Power grid winding side 152: Compensation winding side 153: Load winding side 160: Grid-connected circuit breaker 170: High-voltage side of the power grid 180: Low-voltage distribution radio area 10: Physical Topology Layer 20: Long-term scale optimization scheduling layer 30: Real-time control layer on the energy storage side 40: Real-time control layer on the network side 50: Real-time control layer on the network side 301: DC bus PI controller 302: Low-pass filter module 303: SOC Status Monitoring Module 304: Supercapacitor Power Limiting Module 305: Battery power limiting module 400: Network-side controller 401: Power Calculation Module 402: VSG control loop 403: Voltage outer loop 404: Inner Current Loop 405: SVPWM module 500: Network-side controller 501: Phase-Locked Loop 502: Multidimensional Disturbance Sensing and Compensation Loop 503: Sequence Separation and Harmonic Detection Module 504: Compensation Current Calculation and Vector Superposition Module 505: Reference Power Control Loop 507: Local Transient Power Sensing and High-Frequency Smoothing Loop 508: Power Deviation Calculation Module 509: High-pass filter module 510: Current Smoothing Calculation Module 511: Instruction Synthesis Module 512: Phase-by-phase current safety clamping module 513: Current Inner Loop Controller 514: SVPWM module Detailed Implementation

[0027] To make the objectives, technical solutions, and advantages of this invention clearer, the technical solutions of this invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some, not all, of the embodiments of this invention. All other embodiments obtained by those skilled in the art based on the embodiments of this invention without creative effort are within the scope of protection of this invention.

[0028] This invention provides a power quality full-function control device. Figure 1 This is a schematic diagram of the multi-time-scale control hierarchy of the power supply quality full-function control device for the distribution network of the present invention. Figure 1As shown, the power quality full-function control device 100 includes: a grid-connected converter 120, a grid-connected converter 110, a hybrid energy storage system 130, a three-winding transformer 150, and a controller; the grid-connected converter 120 and the grid-connected converter 110 share a DC bus 140, and the hybrid energy storage system 130 is connected to the DC bus 140; the AC side of the grid-connected converter 120 is connected to the load winding side 153 of the three-winding transformer 150, and the AC side of the grid-connected converter 110 is connected to the compensation winding side 152 of the three-winding transformer 150; the grid winding side 151 of the three-winding transformer 150 is connected to the upstream power grid (high-voltage side 170 of the power grid) via a grid-connected circuit breaker 160; the controller is signal-coupled with the grid-connected converter 120, the grid-connected converter 110, and the hybrid energy storage system 130.

[0029] The controller can be one or more physically, and can be either a centralized controller or multiple distributed controllers. For example, the controller may include a grid-side controller 400 and a grid-connected controller 500, which respectively undertake the control functions of the grid-connected converter and the grid-connected converter. The present invention does not limit the number of controllers.

[0030] refer to Figure 1 The low-voltage distribution substation 180 is connected to distributed energy storage and distributed power sources such as photovoltaics and wind turbines; the high-voltage side 170 of the power grid is connected to the grid winding side 151 of the three-winding transformer 150 via the grid-connected circuit breaker 160 to connect to the low-voltage distribution substation 180. The hybrid energy storage system 130 adopts a composite architecture including energy-type energy storage unit 131 (battery) and power-type energy storage unit 132 (supercapacitor), and is connected to the common DC bus 140 side of the grid-connected converter 120 and the grid-connected converter 110 via the DC / DC converter 133.

[0031] In some embodiments, the power quality full-function control device 100 adopts a back-to-back power electronic converter topology with a common DC bus 140. The AC side of the grid converter 120 is connected to the load winding side 153 of the three-winding transformer 150 via an LCL filter 121. The AC side of the grid converter 110 is connected to the compensation winding side 152 of the three-winding transformer 150 via an LC filter 111. The hybrid energy storage system 130 includes a power-type energy storage unit 132 and an energy-type energy storage unit 131, which are connected to the DC bus 140 via a DC / DC converter 133.

[0032] Specifically, Figure 2 This is a schematic diagram of the back-to-back converter main circuit topology of the power quality full-function control device provided by this invention. Figure 2As shown, both the grid-connected converter 120 and the grid-connected converter 110 adopt a midpoint clamped three-level topology. A common DC bus 140 and a neutral point are formed by a first capacitor 141 and a second capacitor 142 connected in series on the DC side. The LCL-type filter 121, composed of an inverter-side inductor 121B, a grid-side inductor 121C, and a first filter capacitor 121A, is used to filter out high-frequency switching harmonics and ensure grid-connected current quality. The LC-type filter 111, composed of a second filter capacitor 111A and a filter inductor 111B, is used to establish a low-distortion voltage waveform. The midpoint clamped three-level topology is used to reduce the total harmonic distortion of the output voltage and current, as well as the voltage stress on the switching devices. The common DC bus 140 is used to achieve power decoupling of the AC systems on both sides and provides a hardware channel for energy dispatching and dynamic power mutual assistance at multiple time scales.

[0033] This invention also provides a multi-time-scale optimization control method applied to the aforementioned full-function power quality control device 100. This multi-time-scale optimization control method is executed by a controller and includes: a normal grid-connected operation mode and a fault-off-grid operation mode.

[0034] In the normal grid-connected operation mode, a two-layer collaborative control based on decoupling of long and short time scales is executed: On the long time scale, global optimization scheduling aimed at distribution area energy management and peak shaving and valley filling is performed, including: solving for a steady-state reference operation command set based on a pre-built global optimization scheduling model containing a dynamic capacity reservation mechanism; wherein, the dynamic capacity reservation mechanism is used to limit the upper limit of the output of the grid-type converter 110 and the grid-following converter 120 in positive-sequence fundamental wave energy scheduling, so as to forcibly reserve dynamic capacity margin for transient inertia response and power quality compensation; on the short time scale, the grid-type converter 110 is controlled based on a virtual synchronous generator mechanism to adjust its output according to different operating conditions to support the voltage quality of the distribution area; while controlling the grid-following converter 120 to execute the steady-state reference operation command set, the current of the grid-following converter 120 is adjusted to perform comprehensive power quality compensation locally;

[0035] In the fault-off-grid operation mode, time-series coordinated control guided by ensuring system lifetime is executed: when a grid anomaly is detected, the grid-connected circuit breaker 160 is disconnected, wherein the global optimized scheduling is correspondingly suspended, and the distribution substation 180 constitutes an islanded microgrid; during the transient phase of the islanded microgrid, the grid-type converter 110 is controlled to smoothly switch to the dominant voltage source, and the amplitude, phase, and frequency reference of the islanded microgrid bus voltage is established; the grid-following converter 120 is controlled to lock the phase and frequency of the islanded microgrid bus voltage in real time through a phase-locked loop, while simultaneously acquiring and tracking its amplitude; after the islanded microgrid enters a steady state, based on the pre-constructed islanded coordinated optimized scheduling model and the real-time status of the hybrid energy storage system 130, graded load shedding protection is executed; when the upstream grid recovers to a preset qualified state and meets the synchronous parallel conditions, the grid-type converter 110 and the grid-following converter 120 are controlled to perform grid reset and reconnect to the global optimized scheduling.

[0036] The power quality full-function control device and its multi-time-scale optimization control method provided by this invention have the following technical effects: 1. A two-layer collaborative control architecture with decoupled long and short time scales is constructed. The upper layer uses global optimization scheduling on a long time scale to formulate a baseline operating strategy to achieve efficient energy management and peak shaving in the distribution area. Combined with real-time disturbance perception and dynamic compensation on a short time scale at the lower layer, transient support and local power quality management are achieved. This architecture breaks through the limitations of traditional single-point management and single-time-dimensional control of converters, realizing the physical separation and logical unification of macro-energy scheduling and micro-power quality management. It significantly improves the operational stability of active distribution networks under varying operating conditions and substantially reshapes the impedance characteristics of distribution nodes, enhancing the support rigidity of local weak power grids. 2. Full-condition system-level multi-objective collaborative control of grid-connected converters, grid-linked converters, and their downstream loads is achieved. Under normal grid-connected conditions, the potential of the grid-connected and grid-connected converters is fully utilized, with no idle converters. Through the coordinated operation of the grid-connected and grid-connected converters, high-quality power supply to the load is achieved. Under grid fault off-grid conditions, the grid-connected converters establish an island voltage reference, and the grid-connected converters maintain continuous tracking of the bus voltage, ensuring the transient stability of the islanded microgrid and the highly reliable and continuous power supply to critical and sensitive loads.

[0037] In the fault-off-grid operation mode, the present invention adopts a four-stage time-series coordinated control strategy guided by the lifespan: (1) Fast transient isolation and electrical topology reconfiguration: When a short-circuit fault or severe voltage quality degradation is detected in the upstream grid, the grid connection point is disconnected; the grid-type converter maintains the core equation of the virtual synchronous machine unchanged and transitions to the dominant voltage source to establish the island voltage and frequency reference; the grid-following converter maintains the reference command before the fault and continuously and smoothly tracks the local bus voltage. (2) Adaptive stability control: Extract the angular frequency deviation and angular frequency change rate of the island microgrid, and construct a phase trajectory region discrimination mechanism based on the sign of their product. Increase the virtual moment of inertia during the frequency acceleration deviation stage, and reduce the virtual moment of inertia back to the reference value during the frequency recovery stage, and superimpose a composite damping compensation term based on the absolute value of the frequency deviation and the absolute value of the rate of change to suppress frequency recovery overshoot and accelerate transient convergence. (3) Islanding steady-state collaborative optimization and graded load shedding protection: After the islanding enters steady state, an islanding collaborative optimization scheduling model is established with the goal of minimizing voltage deviation, load shedding penalty and optimizing state of charge; the grid-connected converter is controlled as a controlled power node to bear the basic load, and the grid-connected converter is passively supplemented for power difference as an islanding balancing node. Combining the optimized scheduling power allocation results and real-time state of charge, the expected maximum power support time is recursively calculated. When the time safety threshold or power warning threshold is reached, graded load shedding logic is executed according to the priority of load importance. (4) Seamless grid-connected reset based on pre-synchronization compensation: After the upstream grid is restored, the grid-connected converter activates the pre-synchronization compensation control loop to eliminate deviation; after the synchronization condition is met, the circuit breaker is closed, and the grid-connected converter and the grid-connected converter reset their states and reconnect to the long-term global optimization scheduling framework.

[0038] In some embodiments, constructing a global optimization scheduling model including a dynamic capacity reservation mechanism specifically includes: establishing a comprehensive objective function, wherein the comprehensive objective function takes minimizing steady-state voltage deviation, optimizing the power factor at the point of common coupling, reducing system active power losses, and optimizing the state of charge of the hybrid energy storage system 130 as joint optimization objectives; establishing system steady-state operation constraints, wherein the system steady-state operation constraints include: setting the upper limit coefficient of steady-state capacity utilization of the grid-connected converter 110 and the upper limit coefficient of symmetrical fundamental frequency capacity utilization of the grid-connected converter 120 respectively; and constructing the global optimization scheduling model based on the comprehensive objective function and the system steady-state operation constraints.

[0039] The above embodiments have the following technical effects: improving the economy and power supply quality of the system in steady state operation, ensuring that the control device always has a DC energy support baseline to cope with sudden transient faults; ensuring that grid-type converters and grid-connected converters always retain dynamic capacity margin, avoiding single-phase current overload and hardware blockage.

[0040] In some embodiments, controlling the grid-connected converter 110 to adjust its output based on a virtual synchronous generator mechanism according to different operating conditions in a short timescale to support the voltage quality of the distribution substation specifically includes: when the distribution substation 180 is under heavy overload and the steady-state output of the grid-connected converter 120 is limited, based on the virtual synchronous generator mechanism and combined with the real-time capacity status of the grid side, locally correcting the steady-state reference operating command set, controlling the grid-connected converter 110 to inject steady-state active current across ports to compensate for the three-winding transformer... The voltage on the load winding side of transformer 150 drops; when the distribution substation 180 is under non-heavy load and the upstream grid voltage is distorted, based on the virtual synchronous generator mechanism, the negative sequence and harmonic distortion components of the upstream grid voltage are extracted, and the grid-type converter 110 is controlled to superimpose the corresponding compensation voltage to perform bus voltage waveform repair; when the upstream grid experiences transient disturbances, based on the virtual synchronous generator mechanism, the dynamic capacity margin is invoked, and the grid-type converter 110 is controlled to autonomously output transient power compensation in response to the frequency and voltage changes of the upstream grid.

[0041] The above embodiments have the following technical effects: they indirectly reduce the equivalent internal resistance of the system, improve the rigidity and short-circuit capacity of the power grid, substantially reshape the impedance characteristics of the distribution nodes, improve the support rigidity of the local weak power grid, and ensure the purity of the power supply waveform of local critical loads from the voltage dimension.

[0042] In some embodiments, while controlling the grid-connected converter 120 to execute the steady-state reference operation command set on a short time scale, adjusting the current of the grid-connected converter 120 to perform comprehensive power quality compensation locally includes: extracting the negative sequence current component, harmonic current component, and local high-frequency power deviation component from the load winding side of the three-winding transformer 150, and performing vector synthesis with the steady-state reference operation command set to generate a comprehensive reference command; applying a phase-segment current safety clamping mechanism to the comprehensive reference command, prioritizing closed-loop tracking of the steady-state reference active power command when the single-phase current approaches the physical safety limit, and dynamically reducing the amplitude of the power quality compensation command, so as to adjust the output current of the grid-connected converter 120 in a closed loop to perform comprehensive power quality compensation locally. The comprehensive power quality compensation includes: performing high-frequency smoothing of three-phase imbalance, waveform distortion, and local transient fluctuations. The above embodiments have the following technical effects: ensuring that the grid-connected converter 120 does not disconnect from the grid due to single-phase overload.

[0043] In the above embodiments of the present invention, dynamic coordination of the imbalance and harmonic comprehensive flexible management of the grid-connected converter, and the additional output of the grid-connected converter under heavy overload, form a strong support for the grid-connected converter.

[0044] In some embodiments, the method further includes: adjusting the virtual moment of inertia and damping coefficient of the grid-type converter 110 based on the angular frequency deviation and rate of change of the islanded microgrid, specifically including: extracting the angular frequency deviation and rate of change of the islanded microgrid in real time; dynamically adjusting the virtual moment of inertia of the grid-type converter 110 according to the product of the angular frequency deviation and the rate of change of the angular frequency: when the product is greater than zero, increasing the virtual moment of inertia of the grid-type converter 110; when the product is less than or equal to zero, reducing the dynamic virtual moment of inertia to a preset steady-state baseline value; and dynamically adjusting the damping coefficient of the grid-type converter 110 based on the absolute value of the angular frequency deviation and the absolute value of the rate of change of the angular frequency.

[0045] The above embodiments have the following technical effects: dynamically increasing the virtual moment of inertia to absorb transient impact energy, reducing the virtual moment of inertia back to the steady-state basic virtual moment of inertia to eliminate the inertial drag effect and suppress the overshoot phenomenon in the frequency recovery stage, and superimposing composite damping compensation to provide dynamic damping dissipation force to accelerate island transient convergence.

[0046] In some embodiments, constructing an islanded collaborative optimization scheduling model specifically includes: establishing an objective function for the islanded microgrid, wherein the objective function takes minimizing steady-state voltage deviation, minimizing tiered load shedding penalties, and optimizing the state of charge of the hybrid energy storage system 130 as joint optimization objectives, and the tiered load shedding penalties introduce node load importance penalty coefficients and shedding ratio optimization variables; reconstructing the constraints of the islanded microgrid: configuring the grid-connected converter 120 and distributed power sources as controlled power nodes to bear the islanded basic load, configuring the grid-connected converter 110 as the islanded balancing node to make up for the power difference, and configuring the hybrid energy storage system 130 to maintain the voltage stability of the DC bus 140; and constructing the islanded collaborative optimization scheduling model based on the objective function and the constraints.

[0047] The above embodiments have the following technical effects: reducing the steady-state load rate of the grid-type converter 110 and avoiding triggering its hardware protection limit.

[0048] In some embodiments, after the islanded microgrid enters a steady state, based on a pre-built islanded collaborative optimization scheduling model and the real-time status of the hybrid energy storage system 130, tiered load shedding protection is executed. Specifically, this includes: periodically calculating the expected maximum power support time of the islanded microgrid based on the real-time state of charge, safe state of charge limit, rated total energy storage capacity, and current filtered steady-state discharge active power of the energy storage units 131 in the hybrid energy storage system 130; and updating the islanded load importance priority list according to the solution results of the islanded collaborative optimization scheduling model; when the expected maximum power support time is lower than a preset time safety threshold, or the real-time state of charge is lower than a preset power warning threshold, tiered load shedding protection is triggered, and non-critical load nodes are disconnected in the order of the islanded load importance priority list.

[0049] The above embodiments have the following technical effects: by implementing graded load shedding protection control, the continuous power supply to critical loads is ensured.

[0050] In some embodiments, when the upstream power grid recovers to a preset qualified state and meets the synchronization and paralleling conditions, the grid-connected converter 110 and the grid-linked converter 120 are controlled to perform grid-connection reset and reconnect to the global optimized scheduling. Specifically, this includes: before closing the grid-connected circuit breaker 160, activating the pre-synchronization compensation control loop of the grid-connected converter 110 (see...). Figure 1 In the pre-synchronous compensation stage, the phase and amplitude compensation commands between the upstream power grid and the islanded microgrid are used as feedforward fine-tuning signals and superimposed on the virtual rotor motion equation and excitation control equation to eliminate deviations. After the synchronous parallel conditions are met and the grid-connected circuit breaker 160 is closed, the grid-connected converter 110 is controlled to reset the steady-state reference command set and redetermine the dynamic capacity margin. The grid-connected converter 120 is controlled to return to the controlled current source mode and reconnect to the global optimized scheduling.

[0051] The above embodiments have the following technical effects: they complete the seamless switching of the distribution radio area 180 from island mode to grid-connected mode, and realize the closed-loop stable operation of the control device 100 under all working conditions of normal grid connection and fault disconnection.

[0052] In some embodiments, the controller in the power quality full-function control device is configured to perform the multi-time-scale optimization control method in any of the above embodiments.

[0053] Figure 6 This is a flowchart illustrating the multi-time-scale optimization control method of the power quality full-function regulation device provided by the present invention. (For example...) Figure 6 As shown, the multi-time-scale optimization control method of the power quality full-function control device 100 includes: normal grid-connected operation mode S100, fault off-grid operation mode S200, and seamless grid-connected recovery phase S300.

[0054] Under normal grid-connected operation mode S100, the following steps are performed:

[0055] Step S101: The long-term scale optimization scheduling layer 20 acquires network parameters and multi-time scale operation data, establishes a long-term scale global optimization scheduling model including two-dimensional dynamic capacity reservation constraints, and solves to generate steady-state baseline instructions.

[0056] Step S102: The grid-type converter 110, based on the virtual rotor and excitation equations, provides cross-port auxiliary load heavy load power support and calls the reserved margin to provide transient damping and inertia support, thereby improving grid stability; when the load is not under heavy load, it outputs the negative sequence voltage and harmonic compensation of the upstream grid to perform voltage repair function.

[0057] Step S103: Extract negative sequence, harmonic components and high-frequency power deviation from the grid converter 120, and perform multi-target command vector synthesis under the safety clamping mechanism to locally suppress three-phase imbalance, waveform distortion and transient fluctuations.

[0058] Step S104: Maintain the energy balance of DC bus 140, perform hybrid energy storage adaptive allocation based on spectrum layering, allocate high-frequency power to power-type energy storage units, allocate low-frequency power to energy-type energy storage units, and combine initial state of charge (SOC) limiting correction.

[0059] In the S200 fault-free operation mode, the following steps are performed:

[0060] Step S201: After detecting an upstream fault, disconnect the grid connection point (grid circuit breaker 160); the grid-type converter 110 smoothly transitions to establish an island reference as the dominant voltage source, and the phase-locked loop 501 of the grid-type converter 120 smoothly switches to the microgrid bus voltage.

[0061] Step S202: The grid-side controller 400 extracts the frequency deviation and rate of change, adaptively adjusts the dynamic virtual inertia of the grid-type converter 110 based on the phase trajectory region discrimination mechanism, and superimposes composite damping compensation to suppress frequency overshoot.

[0062] Step S203: Construct an islanded collaborative optimization objective function that includes graded load shedding penalties, update power balance constraints, solve the problem by the long-term optimization scheduling layer 20, and issue islanded steady-state scheduling instructions.

[0063] Step S204: Combine the SOC of the hybrid energy storage system 130 to predict the maximum power support time; when the time or SOC reaches the safety threshold, automatically execute graded load shedding protection according to priority to ensure power supply to critical loads.

[0064] During the seamless grid connection recovery phase of S300, the following steps are performed:

[0065] Step S301: After the power grid is restored, the grid-type converter 110 activates the internal pre-synchronization compensation control loop to eliminate the deviation; after the synchronization and parallel operation conditions are met, the switch is closed, each converter is reset and reconnected to the global optimization scheduling framework.

[0066] The above control method will be explained in detail below with specific steps.

[0067] I. Normal Grid-Connected Operation Mode

[0068] When the line is in normal grid-connected operation state S100, the control system adopts a two-layer cooperative control strategy based on multi-time-scale decoupling. In the long-time scale (minute to hour level) optimized scheduling layer 20, energy management and peak shaving of the low-voltage distribution substation 180 are the main focus, priority is given to scheduling the grid-connected converter 120 for steady-state power allocation, and asymmetric steady-state scheduling with dynamic capacity reservation mechanism is implemented for the grid-connected converter 110; in the short-time scale (millisecond to second level) real-time control layer (energy storage side real-time control layer 30, grid-connected side real-time control layer 40), dynamic compensation guided by transient disturbance suppression is performed, and the specific steps are as follows:

[0069] in, Figure 3 This is a schematic block diagram of the hybrid energy storage adaptive power allocation provided by the present invention; Figure 4 This is a schematic block diagram of the virtual synchronous generator of the grid-type converter provided by the present invention; Figure 5 This is a schematic block diagram of the grid-type converter provided by the present invention.

[0070] Step 1: Obtain the basic network parameters and multi-timescale operation data of low-voltage distribution area 180, and construct the source-load state boundary for global optimization scheduling.

[0071] Specifically, the basic network parameters include, but are not limited to, the network topology of the transformer substation (see...). Figure 1 The physical topology layer 10), line impedance parameters, and physical constraint information such as transformer rated capacity limits are included. The multi-timescale operating data includes day-ahead and intraday load forecasts for each node distributed within the low-voltage distribution substation area 180, distributed power generation output forecast curves, and flexible and adjustable resource parameters.

[0072] Preferably, the flexible and adjustable resource parameters include at least the rated capacity, charge and discharge power limit, and initial state of charge (SOC) of the hybrid energy storage system 130; and the rated apparent power limit and maximum permissible overload factor of the grid-connected converter 110 and the grid-connected converter 120.

[0073] Step 2 (corresponding) Figure 6(Middle step S101): Establish a global optimization scheduling model for the distribution network with a dynamic capacity reservation mechanism on a long time scale.

[0074] Specifically, such as Figure 1 As shown in the medium-to-long-term time-scale optimization scheduling layer 20, the global optimization scheduling model aims to obtain the steady-state reference operating instruction set for each scheduling period through day-ahead and intraday long-term time-scale optimization. The reference operating instruction set includes at least: a reference active power reference instruction and a reactive power reference instruction issued to the grid-connected converter 120. The steady-state reference active power setpoint and reactive power setpoint are sent to the grid-type converter 110. The charging and discharging power of each unit in the hybrid energy storage system 130, as well as the active and reactive power dispatching commands of each distributed power source and distributed energy storage in the low-voltage distribution area 180.

[0075] The global optimization scheduling model includes a long-term energy management objective function and system steady-state operation constraints:

[0076] (1) Objective function: Establish a comprehensive objective function that includes steady-state voltage deviation, power factor at point of common coupling, system active power loss, and hybrid energy storage SOC:

[0077]

[0078]

[0079]

[0080]

[0081]

[0082] In the formula, , , and These respectively represent the weight coefficients of the corresponding objective sub-functions; This represents the total number of time periods within the scheduling cycle. For distribution network nodes; For nodes exist The steady-state voltage amplitude at time t; The node's rated voltage; for The reactive power flowing out of the main transformer outlet at any given moment; For the collection of distribution network branches, for Flowing through the side road The current amplitude, branch road The resistance; and For batteries and supercapacitors in hybrid energy storage systems Real-time SOC at any given moment; and These are the optimal SOC reference values ​​for ensuring energy mutual support across multiple time scales. Preferably, the reference values ​​are dynamically set based on the system's real-time control tasks. For example, during off-peak periods, a high reference value is set to achieve off-peak charging reserves, while during high-load periods, the reference value is dynamically lowered to release capacity to support load power demand.

[0083] Steady-state voltage deviation evaluation index The aim is to suppress long-term node voltage exceedances in distribution transformer areas caused by distributed power source access or load peak-valley variations, thereby improving the steady-state voltage power supply quality; the power factor penalty index at the point of common coupling. The aim is to dispatch grid-connected converters to provide steady-state reactive power compensation, reducing the reactive power load on the upstream main transformer; system active power loss assessment indicators. The aim is to improve the economic operation of the system by optimizing the power flow distribution within the distribution area; SOC offset penalty index The aim is to prevent deep charging and discharging of energy storage units and ensure that the control device always has a DC energy support baseline to cope with sudden transient faults.

[0084] (2) System steady-state operation constraints: The operation constraints include basic network operation constraints and internal power coupling and capacity reservation constraints of the power quality full-function control device 100. The specific constraints are as follows:

[0085] a) Safety constraints on node voltage and branch current:

[0086]

[0087] In the formula, and For nodes The upper limit of voltage fluctuation; branch road The overload critical current.

[0088] b) Transformer and distributed power generation capacity constraints:

[0089]

[0090] In the formula, and for There is reactive power flowing out of the main transformer outlet at all times; , , and These are the upper and lower limits of the active and reactive power output of the main transformer, respectively. and for The active and reactive power generated by distributed generation (DG) at all times. This is the day-ahead forecast of the active power generated by DG. and These represent the upper limit of reactive power output of DG.

[0091] c) Operational constraints of distributed energy storage:

[0092]

[0093] In the formula, yes The energy storage capacity of a distributed energy storage system (DESS) at any time. and for The active power of DESS during charging and discharging at any given time. and The charge / discharge efficiency of DESS. This is the rated energy storage capacity of DESS. and These are the upper and lower limits of the state of charge of DESS, respectively. This is the rated active power of the DESS.

[0094] d) System power flow constraints:

[0095]

[0096] In the formula, the subscript , or Indicates the node number of the distribution network; and Representing the nodes respectively A set of connected child and parent nodes; , and These are the active and reactive power at the beginning of the line and the branch lines. The reactance; and It is a node The total active and reactive power output of all power supplies on the device. and It is a node The total active and reactive power consumed by the load.

[0097] e) Hybrid energy storage system 130 operational constraints:

[0098] As a trans-timescale energy pool for the control device, its charging and discharging power and state of charge must meet strict physical boundaries. These constraints need to be defined independently for both the energy storage unit 131 (battery) and the power storage unit 132 (supercapacitor). Taking the battery as an example, its physical constraint boundaries are:

[0099]

[0100] In the formula, yes The actual stored energy of the battery at all times; and Each of them Active power during charging and discharging at any given time; and Their respective charge and discharge efficiencies are... Its rated energy storage capacity; and These are the lower and upper limits of its fully charged state, respectively; and These are the maximum charge / discharge power limits, respectively. Similarly, the energy state model and basic charge / discharge constraints of the supercapacitor are similar to those of the battery. Optionally, considering the significant differences in the physical response characteristics of the two types of energy storage media, independent ramp rate constraints can be added to each to finely limit the impact of drastic power changes on the lifespan of the energy storage media.

[0101] f) Power coupling constraints on the common DC bus 140:

[0102] Since the grid-connected converter 110 and the grid-connected converter 120 share the DC bus 140, their AC / DC energy conversion must be dynamically balanced through the hybrid energy storage system 130. Ignoring internal switching losses in the converter, its DC-side power coupling equation is:

[0103]

[0104] In the formula, the reference power command specifies that the outflow DC bus 140 is positive. This constraint enables cross-port power mutual assistance of the device under steady-state scheduling.

[0105] g) Asymmetric scheduling and dynamic capacity reservation constraints for multi-functional converters:

[0106] To ensure the transient support capability of the control device 100 under sudden disturbances, an asymmetric capacity constraint mechanism is implemented for the grid-type converter 110 and the grid-connected converter 120:

[0107]

[0108] In the formula, and These represent the rated apparent power of the two types of converters, respectively.

[0109] In particular, ( The coefficient for steady-state capacity utilization of the grid-connected converter 110 is defined as the upper limit coefficient. This constraint physically limits its steady-state output during normal grid connection, ensuring that it always retains at least [a certain value]. Its dynamic current margin is specifically designed for responding to short-timescale transient virtual inertia support and seamless fault switching. ( The coefficient for the symmetrical fundamental frequency capacity utilization of the grid-connected converter 120 is preferably set based on the historical maximum imbalance of the distribution substation area. Value (e.g., set to 0.8). By setting this constraint, its output limit in positive-sequence fundamental wave energy scheduling is restricted at the global scheduling level, forcing a reserved capacity for it. The asymmetric management and harmonic compensation capacity margin avoid single-phase current overload and hardware blockage when negative sequence compensation commands are superimposed on a short time scale.

[0110] Furthermore, intelligent optimization algorithms, including but not limited to improved particle swarm optimization, differential evolution, or interior-point methods, are used to solve the established long-term global optimization scheduling model for the distribution network. The optimization variables in the solution should include at least the following: , , , , , , , , , , , Once the algorithm meets the convergence condition, it outputs the global optimal solution and... Figure 1 As shown in the information flow, the reference command and set value are sent to each converter and distributed resource.

[0111] Optionally, the model incorporates a capacity assessment mechanism: if the calculated peak power or total energy demand exceeds the current configuration, an early warning and suggested value are added to the output capacity.

[0112] Step 3: Execute multi-objective dynamic cooperative control based on decoupling of long and short time scales.

[0113] Specifically, the converter controllers (grid-side controller 400 and grid-connected controller 500) use the steady-state reference operating command set as the basic operating point, and combine it with locally acquired transient disturbance signals to perform mutually decoupled low-level closed-loop control on the grid-connected converter 110 and the grid-connected converter 120. The basic control logic is as follows: Figure 1 The network-side real-time control layer 40 and the network-side real-time control layer 50 are shown.

[0114] (1) Control the grid-type converter 110 to perform composite voltage source control based on steady-state auxiliary and transient rigid support (corresponding to) Figure 6 (Middle step S102):

[0115] like Figure 1 The real-time control layer 40 on the network side and Figure 4 As shown, the grid-connected converter operates in grid-connected mode as a controlled voltage source. Its internal Virtual Synchronous Generator (VSG) control loop 402 receives the reference active power setpoint. With reactive power setpoint The power calculation module 401 calculates the actual output active power by collecting the voltage and current of the point of common coupling (PCC) in real time. With reactive power And input it into the governing equation (13-14):

[0116] a) Execute the virtual rotor motion equations for active-frequency control:

[0117]

[0118] In the formula, The virtual moment of inertia of the grid-type converter 110; The rated angular frequency of the upstream power grid (170° on the high-voltage side of the power grid); The virtual angular frequency generated for the control loop; The active power-frequency droop factor; This is the virtual damping coefficient.

[0119] b) Execute the excitation control equations for reactive-voltage control:

[0120]

[0121] In the formula, and These are the rated voltage amplitude and the virtual internal potential amplitude generated in real time by the VSG control loop 402, respectively. This is the reactive power-voltage droop coefficient.

[0122] c) Steady-state assist and transient disturbance rejection mechanisms:

[0123] Regarding steady-state load compensation: when the distribution transformer area is operating normally and no overload occurs, the optimized distribution... The rotor motion equation guides the converter into zero-power standby mode; when the load is heavily overloaded and the grid-side capacity is at its peak, it receives... Specifically, due to the lag in upper-level optimization, the grid-connected converter 110 can directly correct the upper-level reference command locally based on the real-time capacity status and actual load status of the grid-connected converter 120. Subsequently, the virtual phase angle inside the converter... The converter actively leads the grid voltage phase angle, outputting steady-state active current. This current is injected into the magnetic circuit through the compensation winding 152, utilizing the magnetic flux superposition effect to compensate for the load voltage drop. Under this operating condition, the grid-type converter 110 is equivalent to a low-impedance controlled voltage source connected in parallel with the upstream grid (high-voltage side 170), which indirectly reduces the equivalent internal resistance of the system, improves the grid rigidity and short-circuit capacity, and enhances its heavy-load resistance without changing the grid frequency.

[0124] Regarding voltage waveform repair: When the system is operating under non-heavy load conditions and the grid-connected converter 110 has sufficient capacity margin, the controller identifies and extracts the negative sequence and harmonic distortion components of the grid-connected point voltage in real time. The grid-connected converter 110 utilizes its controlled voltage source characteristics to superimpose reverse voltage compensation commands into its internal virtual potential, thereby achieving waveform repair and management of upstream grid voltage distortion. This function is decoupled from the current-type management of the grid-connected converter 120, ensuring the purity of the power supply waveform for local critical loads from a voltage perspective.

[0125] Regarding transient disturbance immunity: If a sudden frequency shift or voltage fluctuation occurs in the upstream power grid (high-voltage side of the power grid 170), resulting in... or To compensate for the error, the grid-type converter 110 directly calls the reserved dynamic current margin according to the above control equations, and automatically generates transient power compensation in response to changes in grid frequency and voltage.

[0126] d) Closed-loop synthesis and PWM drive generation:

[0127] Combining upper-level commands and corrections under different operating conditions, the virtual internal potential amplitude is calculated and generated in real time from the virtual rotor motion equation and excitation control equation. and phase angle The three-phase reference voltage reference command is synthesized and input into the dual closed-loop controller consisting of the voltage outer loop 403 and the current inner loop 404. After the inner loop is regulated, the space vector pulse width modulation (SVPWM) module 405 generates a pulse width modulation signal to drive the underlying power switching device.

[0128] (2) such as Figure 1 Real-time control layer 50 on the network side and Figure 5 As shown, the grid-connected converter 120 is controlled to perform composite current source control based on energy dispatch and flexible power quality management (corresponding to...). Figure 6 (Middle step S103)

[0129] a) Steady-state energy dispatch and benchmark tracking:

[0130] The grid-connected converter 120 controller locks the grid frequency through a phase-locked loop 501, and its reference power control loop 505 directly receives and strictly tracks the reference active power reference command obtained from the long-term optimization solution in step 2. With reactive power reference command This stage aims to address the evolution of base load and deterministic power deficits within the distribution substation area, and, in conjunction with the steady-state throughput of the hybrid energy storage system, achieve macroscopic peak shaving and valley filling, as well as on-site optimization of the transformer power factor.

[0131] b) Multi-dimensional disturbance sensing and integrated power quality compensation:

[0132] In the multi-dimensional disturbance sensing and compensation loop 502, the controller acquires the three-phase voltage and current signals of the transformer load side 153 in real time at high frequency. The sequence separation and harmonic detection module 503 employs, but is not limited to, phase-locked loop and sequence separation techniques based on second-order generalized integrators or delayed signal cancellation to decouple the positive and negative sequence components of the non-sinusoidal, unbalanced load current signal. Based on this, it introduces, but is not limited to, harmonic detection algorithms based on multiple synchronous rotating coordinate systems or instantaneous power theory, to extract multi-dimensional features from the non-sinusoidal, unbalanced load current signal. Specifically, it extracts the negative sequence current component and the main characteristic harmonic current components on the load side, and generates a compensation current reference quantity with equal amplitude and opposite phase to the negative sequence disturbance source and harmonic disturbance source. This is then processed by the compensation current calculation and vector superposition module 504 to generate a comprehensive power quality compensation command. .

[0133] c) Local transient power sensing and high-frequency smoothing:

[0134] In the local transient power sensing and high-frequency smoothing loop 507, the power deviation calculation module 508 calculates the actual net load power of the distribution substation in real time and compares it with the day-ahead reference predicted power. The high-pass filter module 509 extracts the local high-frequency power deviation component caused by sudden changes in the output of distributed power sources or impulsive loads. The required smoothing current is obtained by the smoothing current calculation module 510. In particular, to avoid power coupling and control interference between the two types of converters on the AC side, the grid-connected converter 120 only smooths out the power fluctuations of the local source load and does not perform reverse compensation for the transient power increment output by the grid-connected converter 110.

[0135] d) Decoupling multi-target instructions into closed-loop execution:

[0136] The instruction synthesis module 511 synthesizes the reference instruction representing steady-state energy scheduling. Compensation directives representing power quality management And compensation instructions representing transient stabilization Vector synthesis is performed to construct the final operating reference instruction set for the grid converter 120:

[0137]

[0138] In the formula, Represents the combined current vector; This is a vector synthesis function. Furthermore, a phase-splitting current safety clamping module 512 is introduced before drive execution: utilizing reserved... Capacity margins should prioritize including the power quality compensation instructions. If the instantaneous peak value of a single-phase current after synthesis still approaches the physical safety limit of the converter power device... Prioritize the issuance of steady-state reference active power commands, dynamically reduce the amplitude of negative sequence compensation commands, and ensure that the grid-connected converter 120 does not disconnect from the grid due to single-phase overload. Finally, the clamped integrated reference command is input to the current inner loop controller 513 for zero steady-state error tracking, and the drive signal is generated by the SVPWM module 514.

[0139] (3) Implement hybrid energy storage adaptive power allocation based on DC bus energy balance (corresponding to) Figure 6 (Middle step S104):

[0140] The transient and steady-state power throughput caused by the decoupling control of the grid-type converter 110 and the grid-connected converter 120 on the AC side both converge at the common DC bus. To ensure DC bus voltage stability and efficient utilization of hardware resources, such as Figure 3 and Figure 1 As shown in the real-time control layer 30 on the energy storage side, adaptive power allocation for hybrid energy storage based on spectrum layering is performed:

[0141] The DC bus proportional-integral (PI) controller 301 calculates the total power fluctuation deviation that the hybrid energy storage system needs to smooth on the DC side at the current moment. The low-pass filter module 302 is used to decompose it into high-frequency components and low-frequency components.

[0142] Among them, the high-frequency components, such as the electromagnetic transient inertia response of the grid-type converter 110 and the local sudden disturbances smoothed by the grid-type converter 120, are allocated to the supercapacitor unit with extremely fast response and high power density for instantaneous power support; the low-frequency components, such as optimized command tracking and load trend changes, are allocated to the battery unit with high energy density and suitable for long-term operation for steady-state energy absorption or release.

[0143] Furthermore, before the allocation command is issued, the SOC state monitoring module 303 monitors the state of charge construction correction factor in real time. When it detects that the SOC of any unit is approaching the safety upper or lower limit, it uses a nonlinear mapping function to limit or weight the power command to ensure the underlying energy security of the multi-timescale control device.

[0144] Subsequently, the instructions are processed by the supercapacitor power limiting module 304 and the battery power limiting module 305 respectively, and sent to the corresponding supercapacitor DC / DC controller 133A and battery DC / DC controller 133B to achieve adaptive allocation of high and low frequency power.

[0145] II. Off-grid operation mode in case of failure

[0146] like Figure 6 As shown, when a short-circuit fault or severe voltage quality degradation is detected in the upstream power grid (including a drop in three-phase voltage amplitude below a set threshold, a system voltage imbalance exceeding a preset imbalance limit, or a power grid frequency deviating from the safe operating range), the control device 100 immediately switches to the fault-off-grid operation mode S200. The system stops executing long-term global optimization scheduling and switches to control logic oriented towards ensuring island survival time and transient stability. The specific steps are as follows:

[0147] Step 4 (corresponding) Figure 6 (Middle step S201): Perform fast transient isolation and seamless electrical topology reconfiguration.

[0148] Specifically, the control device continuously monitors the voltage amplitude, frequency, and symmetry of the upstream power grid (high-voltage side 170) of the transformer at high frequency. When the real-time monitored value reaches the power grid anomaly judgment criteria (the criteria include at least: the per-unit value of the three-phase voltage amplitude is lower than a preset value such as 0.1pu, the voltage imbalance is higher than a preset value such as 6%, or the power grid frequency deviates from the preset value of the rated frequency such as ±0.5Hz or more), the system triggers the grid-connected circuit breaker 160 to trip, physically isolating the low-voltage distribution transformer area 180 from the upstream power grid (high-voltage side 170) to form an island topology.

[0149] During the transient moment of electrical isolation, the grid-type converter 110 maintains the core control equation of the virtual synchronous generator unchanged, and relies on the instantaneous extreme energy provided by the power-type energy storage unit 132 to naturally maintain the amplitude, frequency and phase of the output voltage, and establishes a stable independent voltage and frequency reference for the islanded microgrid in a very short time.

[0150] Synchronously, the grid-connected converter 120 maintains controlled current source mode operation, keeping the active and reactive power reference commands constant at the moment before the fault. Since the grid-connected converter 110 maintains the phase and amplitude of the bus voltage during electrical isolation, the phase-locked loop of the grid-connected converter does not need to perform object switching, and can still maintain continuous and smooth tracking of the local microgrid bus voltage. Through the above actions, the grid-connected converter 120, as a controlled power injection node, coordinates with other distributed power sources within the low-voltage distribution substation 180 to complete the electrical reconfiguration of the microgrid topology, avoiding phase jumps and circulating current impacts during switching, and maintaining continuous power supply to critical loads within the distribution substation 180.

[0151] Step 5 (corresponding) Figure 6 (Middle step S202): Execute adaptive stability control based on frequency state feedback and phase trajectory discrimination.

[0152] Specifically, in the initial stage of the isolated island transient state, such as Figure 1 As shown in the real-time control layer 40 on the grid side and the off-grid fault information flow, the controller 400 of the grid-connected converter extracts the angular frequency deviation of the PCC voltage in real time. and the rate of change of angular frequency Furthermore, construct based on The phase trajectory region discrimination mechanism of the product sign (see Figure 1 The adaptive stabilization control system based on frequency state feedback and phase trajectory discrimination is used to determine the dynamic virtual moment of inertia. With dynamic damping coefficient Implement asymmetric adaptive compensation adjustment.

[0153] Preferably, the dynamic virtual moment of inertia The adaptive compensation control law satisfies the following piecewise function:

[0154]

[0155] In the formula, The preset steady-state virtual moment of inertia; The inertia adaptive sensitivity coefficient is greater than zero. When the product... At that time, the virtual moment of inertia is dynamically increased to absorb transient impact energy; when the product When the dynamic compensation term is removed, the dynamic virtual moment of inertia falls back to the steady-state basic virtual moment of inertia. This is to eliminate the inertial drag effect and suppress overshoot during the frequency recovery phase.

[0156] Preferably, the dynamic damping coefficient The composite adaptive compensation control law is as follows:

[0157]

[0158] In the formula, The preset steady-state foundation damping coefficient; and These are the frequency deviation damping gain and the frequency change rate damping gain, respectively. The grid-type converter 110 is configured to provide dynamic damping dissipation force based on the composite compensation term to accelerate islanded transient convergence.

[0159] Step 6 (corresponding) Figure 6 (Middle steps S203 and S204): Perform life-cycle-oriented islanded steady-state collaborative optimization and graded load shedding protection.

[0160] Specifically, after the islanded system enters the steady-state operation phase through transient control, the long-term scale optimization scheduling layer 20 stops executing the economic scheduling of the grid connection phase, and instead constructs the islanded cooperative optimization scheduling model according to the following steps:

[0161] a) Islanding Cooperative Optimization Scheduling and Baseline Command Issuance: Under islanding operation conditions, the islanding cooperative optimization scheduling model establishes an islanding objective function with the joint optimization objectives of minimizing steady-state voltage deviation, minimizing load shedding penalty, and optimizing the state of charge of the hybrid energy storage system.

[0162]

[0163] In the formula, , , These are the weight coefficients for the corresponding objective sub-function; in particular, The tiered load shedding penalty index aims to force the system to prioritize shedding of lower-level loads when energy is scarce. Its expression is:

[0164]

[0165] In the formula, This represents the total number of time periods predicted for islanded scheduling. This refers to the set of all load nodes within the distribution area; For nodes exist Forecast active power load demand at any given time; ( ) is a node exist The cutoff ratio at any given time is optimized as a variable. For pre-configured nodes The load importance penalty coefficient (critical loads are given a high penalty coefficient, and non-critical loads are given a low penalty coefficient).

[0166] Simultaneously, the islanded collaborative optimization scheduling model performs topology and boundary updates based on the system steady-state operation constraints described in step 2: removing substation capacity constraints and upstream main grid power flow injection constraints; updating the islanded node power balance constraints to:

[0167]

[0168] In the formula, , This represents the total output of all distributed power sources and distributed energy storage. Subsequently, the long-term optimization scheduling layer 20 uses an intelligent optimization algorithm to solve the model. The optimization variables solved should at least include: the baseline commands issued to the grid-connected converter 120. , Distributed power generation and distributed energy storage dispatch commands , , , And the optimal cut-off ratio for each load node. .

[0169] like Figure 1As shown in the information flow diagram, the optimal instructions obtained from the solution are sent to each execution unit, and the islanded load importance priority list and disconnection ratio are updated synchronously as the basis for the underlying protection triggering strategy. Specifically, the grid-connected converter 120, distributed power sources, and distributed energy storage act as controlled power nodes to bear the basic load of the island; the grid-connected converter 110 acts as the islanded balancing node to maintain the microgrid voltage and frequency reference and passively responds to compensate for power differences; the hybrid energy storage system 130 passively matches the AC / DC energy mutual assistance needs of each converter to maintain the stability of the common DC bus voltage. Through the above coordination, the steady-state load rate of the grid-connected converter 110 is reduced, avoiding its triggering hardware protection limits.

[0170] b) Survival time assessment and graded load shedding control:

[0171] Based on the power allocation results of the islanded collaborative optimization scheduling model and the real-time SOC of the hybrid energy storage system 130, the energy management system in the long-term optimization scheduling layer 20 periodically calculates the expected maximum power support time of the islanded system. Preferably, the formula for calculating the expected maximum power support time is as follows:

[0172]

[0173] In the formula, This refers to the real-time state of charge of the energy storage unit 131 (battery); The preset safe charge state limit; Its rated total energy storage capacity; To consider the combined discharge efficiency of converter losses and battery internal resistance; The current steady-state discharge active power of the battery after processing by a sliding window filter or low-pass filter (when... (At that time, the support time is set to infinity).

[0174] The condition for triggering the graded load shedding protection logic is: when the estimated maximum power support time... Below the preset time safety threshold, or If the power level falls below the preset warning threshold, the underlying controller will immediately disconnect non-critical load nodes according to the most recently updated priority order to ensure continuous power supply to critical loads.

[0175] Step 7 (corresponding) Figure 6 (Middle step S301): Perform seamless grid-connected reset and full-condition closed-loop operation based on pre-synchronization compensation.

[0176] Specifically, during the fault-off-grid operation mode, the status of the upstream power grid (high-voltage side 170) is continuously monitored. When the monitored status meets the preset recovery judgment indicators (including the upstream power grid voltage amplitude recovering to 0.9~1.1 times the rated value, the power grid frequency recovering to the range of 49.5Hz~50.5Hz, the voltage imbalance meeting the standard, and continuous stable operation for 3~5 seconds), the system triggers the pre-synchronization grid connection process:

[0177] a) Pre-synchronization phase tracking and deviation elimination: Before grid connection, the grid-connected converter 110 activates its internal pre-synchronization compensation control loop (see...). Figure 1 (Inter-regional synchronous compensation link). The controller collects the voltage phase data of the upstream power grid in real time. Phase with the bus voltage of the islanded microgrid Phase compensation commands and amplitude compensation commands are generated by comparison. The compensation commands are used as feedforward fine-tuning signals and superimposed on the virtual rotor motion equation and excitation control equation of the grid-type converter 110 to adjust its output voltage amplitude and phase, so as to actively track and eliminate the instantaneous voltage deviation, frequency deviation and phase deviation between the islanded microgrid and the upstream power grid.

[0178] b) Synchronous Closing and System State Reset: After adjusting the islanded microgrid voltage and the main grid voltage to meet the synchronous parallel conditions (i.e., the voltage amplitude difference, frequency difference, and phase difference on both sides converge to the preset safe dead zone range), the system controls the closing of the grid-connected circuit breaker 160. After closing, the control system performs a smooth mode reset action: the grid-connected converter 110 smoothly returns to the controlled voltage source mode of the normal grid-connected stage, the system resets its steady-state reference active power command to zero or the standby setting value, re-establishes and locks the dynamic capacity reserve margin, and continues to provide basic transient inertia support and auxiliary adjustment for the distribution substation; synchronously, since the main grid voltage and islanded voltage have achieved physical synchronization after the grid-connected circuit breaker 160 is closed, the phase-locked loop 501 of the grid-connected converter 120, while maintaining continuous tracking of the local bus voltage, naturally completes the re-locking of the main grid voltage phase and returns to the controlled current source operation mode of the grid-connected stage without disturbance. The grid-connected converter 120 is reconnected to the long-term global optimization scheduling framework of the distribution substation, and resumes its tasks of peak shaving and valley filling, steady-state energy management, and comprehensive power quality management. This enables the seamless switching of the distribution substation 180 from islanded mode to grid-connected mode, and achieves closed-loop stable operation of the control device 100 under all operating conditions, including normal grid connection (S100) and fault disconnection (S200).

[0179] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, and not to limit them; although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features; and these modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A multi-time-scale optimization control method for a power quality full-function control device, wherein the power quality full-function control device comprises: The system comprises a grid-connected converter, a grid-connected converter, a hybrid energy storage system, and a three-winding transformer; the grid-connected converter and the grid-connected converter share a DC bus, and the hybrid energy storage system is connected to the DC bus; the AC side of the grid-connected converter is connected to the load winding side of the three-winding transformer, and the AC side of the grid-connected converter is connected to the compensation winding side of the three-winding transformer; the grid winding side of the three-winding transformer is connected to the upstream power grid via a grid-connected circuit breaker; the system is characterized in that the multi-time-scale optimization control method includes: a normal grid-connected operation mode and a fault-off-grid operation mode; In the normal grid-connected operation mode, a two-layer cooperative control based on decoupling of long and short time scales is executed: on the long time scale, a global optimization scheduling aimed at distribution area energy management and peak shaving and valley filling is executed, including: solving for a set of steady-state reference operation instructions based on a pre-constructed global optimization scheduling model containing a dynamic capacity reservation mechanism; wherein, the dynamic capacity reservation mechanism is used to limit the upper limit of the output of the grid-connected converter and the grid-following converter in the positive sequence fundamental wave energy scheduling, so as to forcibly reserve dynamic capacity margin for transient inertia response and power quality compensation; In a short timescale, the grid-connected converter is controlled based on a virtual synchronous generator mechanism to adjust its output according to different operating conditions to support the voltage quality of the distribution substation. Simultaneously, the grid-connected converter is controlled to execute the steady-state reference operating command set while adjusting its current to perform comprehensive power quality compensation locally. Specifically, this includes: Extract the negative sequence current component, harmonic current component, and local high-frequency power deviation component from the load winding side of the three-winding transformer, and perform vector synthesis with the steady-state reference operation command set to generate a comprehensive reference command. The integrated reference command is applied with a phase current safety clamping mechanism. When the single-phase current approaches the physical safety limit, the closed-loop tracking of the steady-state reference active power command is maintained first, and the amplitude of the power quality compensation command is dynamically reduced. The output current of the grid-connected converter is adjusted in a closed loop to perform integrated power quality compensation locally. The integrated power quality compensation includes: performing high-frequency smoothing of three-phase imbalance, waveform distortion and local transient fluctuations. In the off-grid fault operation mode, time-series coordinated control guided by ensuring system lifetime is executed: when a grid anomaly is detected, the grid-connected circuit breaker is disconnected, wherein the global optimized scheduling is correspondingly suspended, and the distribution substation forms an islanded microgrid; during the transient phase of the islanded microgrid, the grid-type converter is controlled to smoothly switch to the dominant voltage source, and the amplitude, phase, and frequency reference of the islanded microgrid bus voltage is established; the grid-following converter is controlled to lock the phase and frequency of the islanded microgrid bus voltage in real time through a phase-locked loop, while simultaneously acquiring and tracking its amplitude; after the islanded microgrid enters a steady state, based on the pre-constructed islanded coordinated optimized scheduling model and the real-time status of the hybrid energy storage system, graded load shedding protection is executed; when the upstream grid recovers to a preset qualified state and meets the synchronous parallel connection conditions, the grid-type converter and the grid-following converter are controlled to perform grid reset and reconnect to the global optimized scheduling.

2. The multi-time-scale optimization control method of the power supply quality full-function regulation device according to claim 1, characterized in that, Construct a global optimization scheduling model that includes a dynamic capacity reservation mechanism, specifically including: A comprehensive objective function is established, which takes minimizing steady-state voltage deviation, optimizing the power factor at the point of common coupling, reducing system active power losses, and optimizing the state of charge of the hybrid energy storage system as joint optimization objectives. Establish system steady-state operation constraints, which include setting the upper limit coefficient of steady-state capacity utilization of the grid-type converter and the upper limit coefficient of symmetrical fundamental wave capacity utilization of the grid-type converter, respectively. The global optimization scheduling model is constructed based on the comprehensive objective function and the system steady-state operation constraints.

3. The multi-time-scale optimization control method of the power supply quality full-function regulation device according to claim 1, characterized in that, The method of controlling the grid-type converter based on a virtual synchronous generator mechanism on a short time scale to adjust the output of the grid-type converter according to different operating conditions in order to support the voltage quality of the distribution substation includes: When the distribution substation is under heavy overload and the steady-state output of the grid-connected converter is limited, based on the virtual synchronous generator mechanism and combined with the real-time capacity status of the grid side, the steady-state reference operation command set is locally corrected to control the grid-connected converter to inject steady-state active current across the port to compensate for the voltage drop on the load winding side of the three-winding transformer. When the distribution substation is not under heavy load and the upstream grid voltage is distorted, based on the virtual synchronous generator mechanism, the negative sequence and harmonic distortion components of the upstream grid voltage are extracted, and the grid-type converter is controlled to output the corresponding compensation voltage to perform bus voltage waveform repair. When a transient disturbance occurs in the upstream power grid, based on the virtual synchronous generator mechanism, the dynamic capacity margin is invoked to control the grid-type converter to autonomously output transient power compensation in response to frequency and voltage changes in the upstream power grid.

4. The multi-time-scale optimization control method of the power supply quality full-function regulation device according to claim 1, characterized in that, The method further includes: adjusting the virtual moment of inertia and damping coefficient of the grid-type converter based on the angular frequency deviation and rate of change of the islanded microgrid, specifically including: The angular frequency deviation and angular frequency change rate of the isolated microgrid are extracted in real time. The virtual rotational inertia of the grid converter is dynamically adjusted based on the product of the angular frequency deviation and the rate of change of the angular frequency: when the product is greater than zero, the virtual rotational inertia of the grid converter is increased; when the product is less than or equal to zero, the virtual rotational inertia of the grid converter is reduced to a preset steady-state base value. The damping coefficient of the grid converter is dynamically adjusted based on the absolute value of the angular frequency deviation and the absolute value of the angular frequency change rate.

5. The multi-time-scale optimization control method for the power supply quality full-function regulation device according to claim 1, characterized in that, Constructing an islanded collaborative optimization scheduling model specifically includes: An objective function is established for the islanded microgrid, wherein the objective function takes minimizing steady-state voltage deviation, minimizing tiered load shedding penalties, and optimizing the state of charge of the hybrid energy storage system as joint optimization objectives. The tiered load shedding penalties introduce node load importance penalty coefficients and shedding ratio optimization variables. The constraints for reconstructing the islanded microgrid are as follows: the grid-connected converter and distributed power source are configured as controlled power nodes to bear the basic load of the island, the grid-connected converter is configured as the island balancing node to make up for the power difference, and the hybrid energy storage system is configured to maintain the voltage stability of the DC bus. The island collaborative optimization scheduling model is constructed based on the objective function and the constraints.

6. The multi-time-scale optimization control method of the power supply quality full-function regulation device according to claim 1, characterized in that, After the islanded microgrid enters a steady state, based on the pre-built islanded collaborative optimization scheduling model and the real-time status of the hybrid energy storage system, tiered load shedding protection is executed; specifically including: Based on the real-time state of charge, safe state of charge limit, rated total energy storage capacity, and current filtered steady-state discharge active power of the energy storage units in the hybrid energy storage system, the expected maximum power support time of the islanded microgrid is periodically calculated, and the islanded load importance priority list is updated according to the solution results of the islanded collaborative optimization scheduling model. When the expected maximum power support time is lower than the preset time safety threshold, or the real-time state of charge is lower than the preset power warning threshold, the graded load shedding protection is triggered, and non-critical load nodes are disconnected in the order of the islanded load importance priority list.

7. The multi-time-scale optimization control method for the power supply quality full-function regulation device according to claim 1, characterized in that, When the upstream power grid recovers to a preset qualified state and meets the synchronization and parallel connection conditions, the grid-connected converter and the grid-following converter are controlled to perform grid connection reset and reconnect to the global optimized scheduling, specifically including: Before closing the grid-connected circuit breaker, the pre-synchronous compensation control loop of the grid-type converter is activated. The phase and amplitude compensation command between the upstream power grid and the islanded microgrid is used as a feedforward fine-tuning signal and superimposed on the virtual rotor motion equation and excitation control equation to eliminate the deviation. After the synchronous parallel conditions are met and the grid-connected circuit breaker is closed, the grid-type converter is controlled to reset the steady-state reference operation instruction set and redetermine the dynamic capacity margin. The grid-following converter is then controlled to return to the controlled current source mode and reconnect to the global optimized scheduling.

8. A power quality full-function control device, characterized in that, include: The system comprises a grid-connected converter, a grid-connected converter, a hybrid energy storage system, a three-winding transformer, and a controller. The grid-connected converter and the grid-connected converter share a DC bus, and the hybrid energy storage system is connected to the DC bus. The AC side of the grid-connected converter is connected to the load winding side of the three-winding transformer, and the AC side of the grid-connected converter is connected to the compensation winding side of the three-winding transformer. The grid winding side of the three-winding transformer is connected to the upstream power grid via a grid-connected circuit breaker. The controller is signal-coupled with the grid-connected converter, the grid-connected converter, and the hybrid energy storage system. The controller is configured to perform the multi-timescale optimization control method as described in any one of claims 1 to 7.

9. The power quality full-function control device according to claim 8, characterized in that, The power quality full-function control device adopts a back-to-back power electronic converter topology with a common DC bus. The AC side of the grid-connected converter is connected to the load-side winding of the three-winding transformer via an inductor-capacitor-inductor LCL filter. The AC side of the grid-connected converter is connected to the compensation-side winding of the three-winding transformer via an inductor-capacitor LC filter. The hybrid energy storage system includes power-type energy storage units and energy-type energy storage units, which are connected to the DC bus via a DC-DC / DC converter.

Images