Grid-connected and off-grid seamless switching method and system for constructed energy storage micro-grid

Abstract

The invention provides a grid-connected and off-grid seamless switching method and system for a network-constructed energy storage micro-grid, belongs to the technical field of grid-connected and off-grid switching, and realizes rapidity and stability of micro-grid mode switching through multi-stage cooperative control. The method comprises the following steps: a power grid state real-time monitoring stage: acquiring power grid short-circuit capacity, impedance angle, voltage amplitude, frequency and phase parameters through impedance spectrum analysis and phase-locked loop cooperative detection; a pre-synchronization-hybrid control-dynamic damping three-stage switching stage; a steady state maintaining stage: maintaining a system damping ratio gt through a multi-target cooperative controller; and in the whole switching process, the bus voltage fluctuation is less than or equal to 2%, and the frequency deviation is less than or equal to 0.1 Hz. According to the method, MPC pre-synchronization, hybrid control of dynamic weight, virtual impedance adaptive compensation and multi-target collaborative optimization are controlled through model prediction, so that the problems are solved, seamless switching is realized, and the quality of electric energy is improved.

Description

Technical Field

[0001] The present invention relates to the technical field of grid-connected and off-grid switching, and in particular to a method and system for seamless grid-connected and off-grid switching of a grid-connected energy storage microgrid. Background Art

[0002] When switching between grid-connected and off-grid modes, the following problems are likely to occur due to the sudden change in the control characteristics of grid-forming energy storage (Grid-Forming Energy Storage), grid impedance differences, and delayed dynamic response of power balance: (1) Insufficient synchronization accuracy: The traditional phase-locked loop (PLL) has a large phase detection error when the grid is distorted, resulting in switching shock (such as >5% voltage fluctuation); (2) Mode switching oscillation: When switching between voltage source and current source control modes, transient instability is easily caused due to the fixed weight (damping ratio <0.5); (3) Poor impedance adaptability: The virtual impedance parameters are fixed and cannot adapt to the dynamic changes of the main grid impedance (such as SCR=2~8 scenarios), resulting in harmonic resonance; (4) Wind and solar power interference: During the switching process, the wind and solar power are forced to exit the MPPT mode, reducing the utilization rate of renewable energy. Therefore, a method and system for seamless switching between grid-forming energy storage and off-grid microgrid is proposed to solve the above problems. Summary of the Invention

[0003] In view of this, the object of the present invention is to provide a method and system for seamless switching between grid-connected and off-grid energy storage microgrids, so as to at least solve the above problems.

[0004] The technical solution adopted in the first aspect of the present invention is as follows:

[0005] A method for seamless switching between grid-connected and off-grid energy storage microgrids, characterized by comprising:

[0006] Real-time monitoring of power grid status: Through impedance spectrum analysis and phase-locked loop collaborative detection, the short-circuit capacity, impedance angle, voltage amplitude, frequency and phase parameters of the power grid are obtained;

[0007] Pre-synchronization-hybrid control-dynamic damping three-stage switching stage:

[0008] Pre-synchronization phase: When a mode switching requirement is detected, a pre-synchronization mechanism based on model predictive control is activated. Using the short-term power support capability of the microgrid energy storage system as a constraint, the output voltage amplitude, frequency, and phase are continuously optimized to ensure that the error with the target grid parameters meets the synchronization thresholds of phase difference <2°, voltage difference <1%, and frequency difference <0.05Hz.

[0009] Hybrid control stage: A dynamic hybrid control model combining voltage source mode and current source mode is constructed. The output ratio of the two modes is adjusted in real time using a weight factor adaptive algorithm derived from Lyapunov stability theory. The weight factor ranges from [0, 1] and is negatively correlated with the grid short-circuit capacity.

[0010] Dynamic damping stage: Based on the positive / negative sequence impedance compensation output by the virtual impedance observer, negative sequence virtual impedance is injected in off-grid mode to suppress voltage imbalance caused by asymmetric loads, and positive sequence virtual impedance is injected in grid-connected mode to enhance power coupling stability with the main grid;

[0011] Steady-state maintenance stage: The system damping ratio is maintained at >0.8 through a multi-objective collaborative controller, and the bus voltage fluctuation during the entire switching process is ≤2% and the frequency deviation is ≤0.1Hz.

[0012] Furthermore, the specific implementation of the weight factor adaptive algorithm includes:

[0013] Establish a Lyapunov function with the state variables of the hybrid control model as input;

[0014] The optimal solution of the weight factor that makes the derivative of the Lyapunov function negative is solved by the gradient descent method;

[0015] The impedance angle is introduced as a feedback quantity into the weight factor iteration process.

[0016] Furthermore, the impedance correction method of the virtual impedance observer includes:

[0017] The recursive least square method is used to identify the equivalent impedance of the power grid online;

[0018] selectively generating a negative-sequence or positive-sequence virtual impedance component according to the grid-connected / off-grid mode, wherein the magnitude of the negative-sequence component is proportional to the unbalance degree of the asymmetric load;

[0019] Impedance compensation closed-loop control is used to make the output impedance track the target impedance curve.

[0020] Furthermore, the pre-synchronization stage further includes:

[0021] When it is detected that the short-circuit capacity of the main network drops to the threshold, the delayed switching logic is triggered;

[0022] During the delay period, the energy storage system is used to maintain the voltage stability of the microgrid, and MPC pre-synchronization is re-executed after the main grid is restored.

[0023] Furthermore, it is characterized in that when the method is applied in a wind-solar-storage hybrid microgrid, the wind-solar power supply maintains maximum power tracking operation, and the power supply interruption time of the critical load is less than 10ms.

[0024] The technical solution adopted in the second aspect of the present invention is as follows:

[0025] Reference Figure 2 A grid-connected and off-grid seamless switching system for a grid-connected energy storage microgrid is provided, which is used to implement a grid-connected and off-grid seamless switching method for a grid-connected energy storage microgrid mentioned in the first aspect. The system comprises:

[0026] A power grid status monitoring device including an impedance spectrum analyzer and an enhanced phase-locked loop circuit;

[0027] Pre-synchronization control device with built-in MPC algorithm processor with power support constraints;

[0028] A multi-mode cooperative control device, comprising a voltage source mode control unit, a current source mode control unit and a weight factor calculation unit;

[0029] A virtual impedance adaptive device, consisting of an impedance observer, a negative-sequence / positive-sequence impedance generator, and an impedance injection circuit;

[0030] The multi-objective collaborative control device uses FPGA to achieve μs-level response damping ratio closed-loop adjustment.

[0031] Furthermore, in the multi-mode collaborative control device:

[0032] The voltage source mode control unit adopts a voltage-frequency regulator based on droop control;

[0033] The current source mode control unit adopts a power feedforward decoupling controller;

[0034] The weight factor calculation unit realizes the real-time solution of Lyapunov function through hardware accelerator.

[0035] Furthermore, the virtual impedance adaptive device further includes:

[0036] Impedance identification module, using recursive least squares processor;

[0037] Mode selection switch, switching the negative sequence or positive sequence impedance path according to the grid-connected / off-grid status signal;

[0038] The impedance compensation verification circuit is used to detect the matching degree between the actual output impedance and the target impedance.

[0039] Furthermore, the FPGA of the multi-objective collaborative control device is configured as follows:

[0040] Perform parallel computation of the MPC algorithm in the pre-synchronization phase;

[0041] Achieve nanosecond-level updates of weight factors in the hybrid control phase;

[0042] The fast switching of positive / negative sequence impedance is completed in the dynamic damping stage.

[0043] Compared with the prior art, the present invention has the following beneficial effects:

[0044] 1. High-precision synchronization: MPC pre-synchronization controls phase / voltage / frequency errors to <2°, <1%, and <0.05Hz, respectively, and switching power impact to <5% of the rated value. Compared with traditional PLL (error >5°), synchronization time is shortened.

[0045] 2. Enhanced dynamic stability: The Lyapunov weighting factor algorithm enables adaptive adjustment of the hybrid control mode, and the system damping ratio is always >0.8; voltage fluctuations are suppressed to ≤2% (traditional methods are >5%), and frequency deviation is ≤0.1Hz.

[0046] 3. Impedance adaptive compatibility: The virtual impedance observer uses the recursive least squares method to identify the grid impedance online and dynamically compensate for the positive / negative sequence components; the voltage imbalance under asymmetric load is <2%, and the harmonic THD is <3%. BRIEF DESCRIPTION OF THE DRAWINGS

[0047] In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the following briefly introduces the drawings required for use in the description of the embodiments. Obviously, the drawings described below are only preferred embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work.

[0048] Figure 1 The present invention provides a schematic diagram of the overall process of a method for seamless switching of a grid-connected energy storage microgrid to and from the grid, according to an embodiment of the present invention.

[0049] Figure 2 It is a schematic diagram of the overall structure of a grid-connected and off-grid seamless switching system for a grid-connected energy storage microgrid proposed in another embodiment of the present invention. DETAILED DESCRIPTION

[0050] The principles and features of the present invention are described below with reference to the accompanying drawings. The enumerated embodiments are only used to explain the present invention and are not used to limit the scope of the present invention.

[0051] Reference Figure 1 An embodiment of the present invention provides a method for seamless switching between grid-connected and off-grid energy storage microgrids, including:

[0052] Real-time monitoring of power grid status: Through impedance spectrum analysis and phase-locked loop collaborative detection, the short-circuit capacity, impedance angle, voltage amplitude, frequency and phase parameters of the power grid are obtained;

[0053] Pre-synchronization-hybrid control-dynamic damping three-stage switching stage:

[0054] Pre-synchronization phase: When a mode switching requirement is detected, a pre-synchronization mechanism based on model predictive control is activated. Using the short-term power support capability of the microgrid energy storage system as a constraint, the output voltage amplitude, frequency, and phase are continuously optimized to ensure that the error with the target grid parameters meets the synchronization thresholds of phase difference <2°, voltage difference <1%, and frequency difference <0.05Hz.

[0055] Hybrid control stage: A dynamic hybrid control model combining voltage source mode and current source mode is constructed. The output ratio of the two modes is adjusted in real time using a weight factor adaptive algorithm derived from Lyapunov stability theory. The weight factor ranges from [0, 1] and is negatively correlated with the grid short-circuit capacity.

[0056] Dynamic damping stage: Based on the positive / negative sequence impedance compensation output by the virtual impedance observer, negative sequence virtual impedance is injected in off-grid mode to suppress voltage imbalance caused by asymmetric loads, and positive sequence virtual impedance is injected in grid-connected mode to enhance power coupling stability with the main grid;

[0057] Steady-state maintenance stage: The system damping ratio is maintained at >0.8 through a multi-objective collaborative controller, and the bus voltage fluctuation during the entire switching process is ≤2% and the frequency deviation is ≤0.1Hz.

[0058] For example, in the real-time monitoring stage of the power grid status, dual-mode detection can be performed, that is, combining impedance spectrum analysis with an enhanced phase-locked loop to build a multi-dimensional parameter monitoring system. The key parameters extracted include short-circuit capacity (reflecting the strength of the power grid), impedance angle (judging inductive / capacitive characteristics), voltage amplitude / frequency / phase (dynamic tracking of the power grid status); the extracted key parameter data can provide a real-time decision basis for subsequent switching; in the three-stage switching stage of pre-synchronization-hybrid control-dynamic damping, for the pre-synchronization stage, model predictive control can be used, that is, rolling optimization can be performed : With the power support capacity of the energy storage system as a constraint (to avoid overcharging / over-discharging), the optimal voltage output trajectory in the future time domain is calculated through the prediction model, and the synchronization threshold is set: phase difference <2°, voltage difference <1%, frequency difference <0.05Hz, to ensure that the power impact during switching is <5% of the rated value, and to perform delayed switching logic: when the main grid short-circuit capacity is insufficient, the energy storage system maintains the microgrid voltage stability and resynchronizes after the main grid is restored; for the hybrid control stage, dynamic hybrid can be achieved through dual modes, namely voltage source mode: providing voltage / frequency support (droop control) and current source mode Formula: Realize accurate power tracking (feedforward decoupling control), and then adjust the output ratio of the two modes in real time through the weight factor adaptive algorithm derived from Lyapunov stability theory. The weight factor range is [0,1] and is negatively correlated with the short-circuit capacity of the power grid. Negative correlation adjustment: the larger the short-circuit capacity of the power grid, the closer the weight factor is to 1 (current source mode dominates), and vice versa, the proportion of voltage source mode increases; for the dynamic damping stage, virtual impedance adaptive adjustment can be used, that is, through the off-grid mode: inject negative sequence virtual impedance to compensate for the negative sequence caused by asymmetric load Current, suppressing voltage imbalance to <2%, through the grid-connected mode: injecting positive-sequence virtual impedance, enhancing the power coupling stiffness with the main grid, avoiding harmonic resonance, and through closed-loop correction: using the impedance observer to correct the compensation amount in real time, so that the output impedance tracks the preset target curve; for the steady-state maintenance stage, the high damping ratio (>0.8) can achieve multi-objective coordinated control with μs-level response through FPGA, and voltage / frequency fluctuation suppression can meet the IEC61000 electromagnetic compatibility standard by switching the bus voltage fluctuation ≤2% and the frequency deviation ≤0.1Hz throughout the entire process.

[0059] This method constructs a complete seamless switching technology system from detection to execution through multi-dimensional parameter monitoring, model predictive control, hybrid mode dynamic adjustment and virtual impedance adaptive compensation, which significantly improves the operational flexibility and power quality of the microgrid.

[0060] The specific implementation of the weight factor adaptive algorithm includes:

[0061] Establish a Lyapunov function with the state variables of the hybrid control model as input;

[0062] The optimal solution of the weight factor that makes the derivative of the Lyapunov function negative is solved by the gradient descent method;

[0063] The impedance angle is introduced as a feedback quantity into the weight factor iteration process.

[0064] For example, for the construction of the Lyapunov function, first, the algorithm takes the state variables of the hybrid control model (such as voltage amplitude, frequency deviation, phase difference, etc.) as input to construct the Lyapunov energy function. This function is mathematically defined as a quadratic function of the system state deviation from the equilibrium point. Its physical meaning is to characterize the potential energy difference between the current state of the system and the stable target. The function formula is:

[0065]

[0066] Among them, x is the state variable vector, Q is a positive definite symmetric matrix. Through the design of this function, the algorithm transforms the control objective into a problem of making V(x) monotonically decrease over time, thereby ensuring the asymptotic stability of the system.

[0067] For the gradient descent method to optimize the weight factor, based on the Lyapunov function, the algorithm iteratively solves the optimal solution of the weight factor α (value range [0,1]) through the gradient descent method, so that the function derivative Satisfy the negative definite condition (i.e. ), the specific steps include:

[0068] Compute the partial derivatives of the Lyapunov function with respect to the weight factors

[0069] Update α along the negative gradient direction:

[0070]

[0071] Among them, η is the learning rate. By dynamically adjusting η, the convergence speed and stability can be balanced.

[0072] The iteration termination condition is Reaching a preset threshold (such as ∈>0) to ensure that the system energy continues to decay to the stable region.

[0073] For the impedance angle feedback correction mechanism, in order to improve the algorithm's adaptability to the dynamic characteristics of the power grid, the grid impedance angle θ is introduced as a feedback quantity into the weight factor iteration process. The impedance angle reflects the inductive / capacitive characteristics of the power grid, and its change directly affects the stability boundary of the hybrid control mode. The specific implementation method is:

[0074] Obtain the grid impedance angle θ(t) in real time through the impedance spectrum analysis module;

[0075] Design the feedback compensation term Δα = f(θ), for example:

[0076] Δα=K P (θ ref -θ)+K i ∫(θ ref -θ)dt

[0077] Among them, K P , K i is the proportional-integral coefficient, θ ref is the desired impedance angle (e.g. 0° corresponds to pure resistance characteristics);

[0078] Add Δα to the update of the gradient descent method to form a closed-loop correction:

[0079]

[0080] This mechanism enables the weighting factor to quickly respond to changes in grid impedance, for example, automatically reducing the current source mode ratio (α→0) in a weak grid (high inductive reactance) scenario to avoid harmonic oscillations.

[0081] The weight factor adaptive algorithm achieves adaptive and smooth switching of hybrid control modes through parameter tuning and real-time feedback correction guided by stability theory, significantly improving the compatibility of grid-connected energy storage systems with complex grid conditions.

[0082] The impedance correction method of the virtual impedance observer includes:

[0083] The recursive least square method is used to identify the equivalent impedance of the power grid online;

[0084] selectively generating a negative-sequence or positive-sequence virtual impedance component according to the grid-connected / off-grid mode, wherein the magnitude of the negative-sequence component is proportional to the unbalance degree of the asymmetric load;

[0085] Impedance compensation closed-loop control is used to make the output impedance track the target impedance curve.

[0086] For example, for online identification of grid equivalent impedance using recursive least squares method, the instantaneous values ​​of the voltage u(t) and current i(t) at the output of PCS (energy storage converter) can be continuously collected by an observer.

[0087] The grid equivalent impedance parameters are updated in real time through the recursive least squares algorithm:

[0088]

[0089] in is the current observation sequence, K(k) is the dynamic gain matrix;

[0090] The identification period is ≤50μs and can track rapid fluctuations of impedance amplitude of ±10%.

[0091] For the grid-connected mode: when the main grid voltage is detected to be normal, a positive sequence virtual impedance component Z_v+=k·Z_eq is generated;

[0092] The k factor ranges from 0.2 to 0.5 and is used to compensate for line voltage drops and enhance power coupling;

[0093] For off-grid mode: When an asymmetric load is detected (e.g. single-phase load > 30%):

[0094] Z_v-=-m·ΔI_neg Λ2 , where ΔI_neg is the effective value of negative sequence current;

[0095] The voltage THD is suppressed to <3% by injecting negative sequence impedance.

[0096] For closed-loop impedance tracking control:

[0097] Establish the target impedance curve: Z_ref=f(ω)=R_ref+jX_ref(ω)

[0098] R_ref=0.5Ω(when connected to the grid) / 1.2Ω(when off the grid)

[0099] X_ref=ωL, where L=2mH(when connected to the grid) / 5mH(when off the grid)

[0100] Impedance compensation verification circuit execution:

[0101] Real-time calculation of output impedance Z_out = u_LLC / i_LLC (LLC resonant cavity parameters)

[0102] After comparing with Z_ref, PI adjustment value ΔZ=K is generated P (Z_ref-Z_out)+K i ∫(Z_ref-Z_out)dt

[0103] The virtual impedance injection value is dynamically adjusted through PWM modulation to make |Z_out-Z_ref|≤0.1Ω.

[0104] The dynamic impedance adjustment mechanism can effectively solve the transient overshoot and impedance mismatch problems that are prone to occur in traditional virtual synchronous machines during mode switching, allowing the system to maintain stable operation within the range of grid strength changes (SCR = 2 to 8).

[0105] The pre-synchronization phase also includes:

[0106] When it is detected that the short-circuit capacity of the main network drops to the threshold, the delayed switching logic is triggered;

[0107] During the delay period, the energy storage system is used to maintain the voltage stability of the microgrid, and MPC pre-synchronization is re-executed after the main grid is restored.

[0108] For example, when the grid-connected energy storage microgrid switches from an off-grid mode to a grid-connected mode, the following steps are performed to achieve seamless anti-disturbance switching:

[0109] (1) Real-time monitoring of main network status

[0110] Continuously collect main grid short-circuit capacity parameters through grid status monitoring devices (e.g., real-time impedance spectrum analysis shows the short-circuit capacity is 500MVA);

[0111] When it is detected that the short-circuit capacity drops to a preset threshold (for example, 30% of the rated capacity, i.e., 150 MVA), the delayed switching protection logic is triggered.

[0112] (2) Delayed switching control execution

[0113] The system automatically enters the 3-second delay hold state;

[0114] Execution during delay:

[0115] The energy storage system activates emergency power support mode, maintaining bus voltage stability at 120% of rated power.

[0116] The pre-synchronization control device suspends MPC parameter optimization and maintains the current voltage / frequency output;

[0117] The multi-objective cooperative controller temporarily increases the damping ratio to 0.85 to enhance disturbance rejection.

[0118] (3) Main network recovery detection and resynchronization

[0119] When the short-circuit capacity of the main grid is detected to be restored to above the threshold (e.g. 200MVA):

[0120] The MPC pre-synchronization algorithm was reactivated to achieve precise alignment with a phase difference of 1.8°, a voltage difference of 0.8%, and a frequency difference of 0.03Hz within 0.5 seconds.

[0121] The hybrid control device starts the transition process dominated by the voltage source mode, and the weighting factor smoothly transitions from 0.3 to 0.7;

[0122] In the dynamic damping stage, positive sequence virtual impedance (5+j2Ω) is injected to match the equivalent impedance characteristics of the main grid.

[0123] (4) Exception handling extension mechanism

[0124] If the main network continues to deteriorate during the delay period (for example, the short-circuit capacity falls below 50MVA):

[0125] Automatically terminate the grid connection attempt and switch to independent microgrid mode;

[0126] Start the non-critical load unloading procedure (giving priority to ensuring power supply to critical loads);

[0127] Generate fault code 0x3F and upload it to the Energy Management System (EMS);

[0128] The three-level protection mechanism of delayed maintenance, energy storage support, and intelligent resynchronization effectively solves the problem of switching failure caused by instantaneous weakening of the main network.

[0129] When the method is applied in a wind-solar-storage hybrid microgrid, the wind-solar power source maintains maximum power tracking operation, and the power supply interruption time of the critical load is less than 10ms.

[0130] For example, in a wind-solar-storage hybrid microgrid system, photovoltaic arrays and wind turbines operate in maximum power point tracking (MPPT) mode, continuously delivering the maximum renewable energy power to the grid; the energy storage system, as the core device of the grid, is responsible for voltage / frequency support and power balance; critical loads (such as data centers and precision manufacturing equipment) have extremely strict requirements on power interruption time. For the continuous operation mechanism of wind and solar power MPPT, the power feedforward decoupling controller in the multi-mode collaborative control device can be used to decouple the output power of the wind and solar power from the regulation capability of the energy storage system; during the grid-connected / off-grid switching process, the voltage source mode control unit (droop control) dynamically compensates for the system power shortage to ensure that the output voltage / frequency fluctuations of the wind and solar power do not affect its MPPT operation status.

[0131] To achieve a power outage time of <10ms for critical loads: in the pre-synchronization stage: the MPC algorithm with FPGA parallel computing is used to complete the rolling optimization of voltage amplitude, frequency, and phase within μs time, locking the phase difference within the range of <2°, shortening the synchronization time before the physical switch action; in the hybrid control stage: the weight factor derived from the Lyapunov function is calculated in real time by the hardware accelerator, so that the voltage source / current source mode switching time is compressed to less than 1ms; in the dynamic damping stage: the virtual impedance adaptive device completes the positive / negative sequence impedance path switching at the moment of mode switching, and ensures the output impedance matching degree >99% through the impedance compensation verification circuit, thereby suppressing switching transient overvoltage.

[0132] Reference Figure 2 Another embodiment of the present invention provides a grid-connected and off-grid seamless switching system for a grid-connected energy storage microgrid, which is used to implement the method proposed in the above embodiment. The system includes:

[0133] A power grid status monitoring device including an impedance spectrum analyzer and an enhanced phase-locked loop circuit;

[0134] Pre-synchronization control device with built-in MPC algorithm processor with power support constraints;

[0135] A multi-mode cooperative control device, comprising a voltage source mode control unit, a current source mode control unit and a weight factor calculation unit;

[0136] A virtual impedance adaptive device, consisting of an impedance observer, a negative-sequence / positive-sequence impedance generator, and an impedance injection circuit;

[0137] The multi-objective collaborative control device uses FPGA to achieve μs-level response damping ratio closed-loop adjustment.

[0138] In the multi-mode cooperative control device:

[0139] The voltage source mode control unit adopts a voltage-frequency regulator based on droop control;

[0140] The current source mode control unit adopts a power feedforward decoupling controller;

[0141] The weight factor calculation unit realizes the real-time solution of Lyapunov function through hardware accelerator.

[0142] The virtual impedance adaptive device further comprises:

[0143] Impedance identification module, using recursive least squares processor;

[0144] Mode selection switch, switching the negative sequence or positive sequence impedance path according to the grid-connected / off-grid status signal;

[0145] The impedance compensation verification circuit is used to detect the matching degree between the actual output impedance and the target impedance.

[0146] The FPGA of the multi-objective collaborative control device is configured as follows:

[0147] Perform parallel computation of the MPC algorithm in the pre-synchronization phase;

[0148] Achieve nanosecond-level updates of weight factors in the hybrid control phase;

[0149] The fast switching of positive / negative sequence impedance is completed in the dynamic damping stage.

[0150] Exemplarily, the power grid state monitoring device consists of an impedance spectrum analyzer and an enhanced phase-locked loop circuit. The impedance spectrum analyzer can use a high-frequency signal injection method (such as a 10kHz pulse sequence) to calculate the equivalent impedance spectrum of the power grid in real time through fast Fourier transform (FFT), and output parameters such as short-circuit capacity and impedance angle. The sampling period is ≤50μs; the enhanced phase-locked loop circuit is based on a second-order generalized integrator design and can simultaneously track the fundamental and harmonic components of the power grid voltage. The pre-synchronization control device has a built-in model predictive control (MPC) algorithm processor, and the hardware implementation includes a power support constraint module and a rolling optimization engine. The power support constraint module can read the energy storage SOC (state of charge) and short-time power capability (such as 1C discharge) in real time, and use it as the linear constraint condition of the MPC optimization problem. The rolling optimization engine can use FPGA to parallelly calculate the voltage / frequency output trajectory of the next 5 control cycles (1ms per cycle) to ensure that the synchronization error meets the phase difference <2° and the voltage difference <1%. The multi-mode coordinated control device includes a voltage source mode control unit, a current source mode control unit, and a weight factor calculation unit. The voltage source mode control unit employs a droop control strategy to automatically adjust the voltage amplitude (e.g., ±2% adjustment range) and frequency (±0.1Hz) based on load changes. The current source mode control unit achieves precise PQ control with a dynamic response time of <5ms via a power feedforward decoupling controller. The weight factor calculation unit solves the Lyapunov function in real time using a hardware accelerator and updates the weight factor α every 50ns. The virtual impedance adaptive device consists of an impedance observer, a mode selector switch, and an impedance injection circuit. The impedance observer uses a recursive least squares processor to update the grid equivalent impedance parameters every 50μs. The mode selector switch switches the impedance path based on the grid-connected / off-grid signal with a response time of <1μs. The impedance injection circuit generates a virtual impedance component using a PWM modulator, injecting a positive-sequence impedance (e.g., 3+j1Ω) when connected to the grid and a negative-sequence impedance (e.g., 0.5-j0.2Ω) when off-grid. The multi-objective collaborative control device implements high-speed closed-loop regulation based on FPGA. The damping ratio control can dynamically adjust the output damping of the energy storage converter through a PID algorithm with μs-level response to maintain a system damping ratio >0.8. The voltage / frequency protection can monitor bus voltage fluctuations (≤2%) and frequency deviations (≤0.1Hz) in real time, and trigger energy storage forced support when the limit is exceeded.

[0151] The system collaborative workflow includes on-grid to off-grid, and off-grid to on-grid. The on-grid to off-grid transition is as follows:

[0152] The grid status monitoring device detects a main grid power outage (voltage drops below 80%) and sends a switching command;

[0153] The presynchronization control device blocks the MPC output, and the multi-mode cooperative control device switches to the voltage source mode (α = 0);

[0154] The virtual impedance adaptive device injects negative sequence impedance to suppress load imbalance;

[0155] The power supply interruption time for key loads during the entire process is <5ms.

[0156] The specific steps for off-grid to grid connection are:

[0157] The impedance spectrum analyzer confirmed that the main grid had recovered (short-circuit capacity > 150MVA), and MPC pre-synchronization was initiated;

[0158] The weight factor gradually increases from 0 to 0.7, and the hybrid control transitions smoothly;

[0159] The multi-objective controller suppresses switching shocks and the bus voltage fluctuation is less than 1.5%.

[0160] The above embodiments show that the system achieves seamless connection and high-reliability operation of microgrid mode switching through hardware acceleration and algorithm optimization.

[0161] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present invention should be included in the scope of protection of the present invention.

Claims

1. A method for seamless switching between grid-connected and off-grid energy storage microgrids, characterized in that: include: Real-time monitoring of power grid status: Through impedance spectrum analysis and phase-locked loop collaborative detection, the short-circuit capacity, impedance angle, voltage amplitude, frequency and phase parameters of the power grid are obtained; Pre-synchronization-hybrid control-dynamic damping three-stage switching stage: Pre-synchronization phase: When a mode switching requirement is detected, a pre-synchronization mechanism based on model predictive control is activated. Using the short-term power support capability of the microgrid energy storage system as a constraint, the output voltage amplitude, frequency, and phase are continuously optimized to ensure that the error with the target grid parameters meets the synchronization thresholds of phase difference <2°, voltage difference <1%, and frequency difference <0.05Hz. Hybrid control stage: A dynamic hybrid control model combining voltage source mode and current source mode is constructed. The output ratio of the two modes is adjusted in real time using a weight factor adaptive algorithm derived from Lyapunov stability theory. The weight factor ranges from [0, 1] and is negatively correlated with the grid short-circuit capacity. Dynamic damping stage: Based on the positive / negative sequence impedance compensation output by the virtual impedance observer, negative sequence virtual impedance is injected in off-grid mode to suppress voltage imbalance caused by asymmetric loads, and positive sequence virtual impedance is injected in grid-connected mode to enhance power coupling stability with the main grid; Steady-state maintenance stage: The system damping ratio is maintained at >0.8 through a multi-objective collaborative controller, and the bus voltage fluctuation during the entire switching process is ≤2% and the frequency deviation is ≤0.1Hz.

2. The method according to claim 1, characterized in that The specific implementation of the weight factor adaptive algorithm includes: Establish a Lyapunov function with the state variables of the hybrid control model as input; The optimal solution of the weight factor that makes the derivative of the Lyapunov function negative is solved by the gradient descent method; The impedance angle is introduced as a feedback quantity into the weight factor iteration process.

3. The method according to claim 1, characterized in that The impedance correction method of the virtual impedance observer includes: The recursive least square method is used to identify the equivalent impedance of the power grid online; selectively generating a negative-sequence or positive-sequence virtual impedance component according to the grid-connected / off-grid mode, wherein the magnitude of the negative-sequence component is proportional to the unbalance degree of the asymmetric load; Impedance compensation closed-loop control is used to make the output impedance track the target impedance curve.

4. The method according to claim 1, wherein The pre-synchronization phase also includes: When it is detected that the short-circuit capacity of the main network drops to the threshold, the delayed switching logic is triggered; During the delay period, the energy storage system is used to maintain the voltage stability of the microgrid, and MPC pre-synchronization is re-executed after the main grid is restored.

5. The method according to any one of claims 1 to 4, characterized in that: When the method is applied in a wind-solar-storage hybrid microgrid, the wind-solar power source maintains maximum power tracking operation, and the power supply interruption time of the critical load is less than 10ms.

6. A grid-connected and off-grid seamless switching system for energy storage microgrids, used to implement the method described in any one of claims 1 to 5, characterized in that: The system comprises: A power grid status monitoring device including an impedance spectrum analyzer and an enhanced phase-locked loop circuit; Pre-synchronization control device with built-in MPC algorithm processor with power support constraints; A multi-mode cooperative control device, comprising a voltage source mode control unit, a current source mode control unit and a weight factor calculation unit; A virtual impedance adaptive device, consisting of an impedance observer, a negative-sequence / positive-sequence impedance generator, and an impedance injection circuit; The multi-objective collaborative control device uses FPGA to achieve μs-level response damping ratio closed-loop adjustment.

7. The system according to claim 6, characterized in that In the multi-mode cooperative control device: The voltage source mode control unit adopts a voltage-frequency regulator based on droop control; The current source mode control unit adopts a power feedforward decoupling controller; The weight factor calculation unit realizes the real-time solution of Lyapunov function through hardware accelerator.

8. The system according to claim 6, wherein: The virtual impedance adaptive device further comprises: Impedance identification module, using recursive least squares processor; Mode selection switch, switching the negative sequence or positive sequence impedance path according to the grid-connected / off-grid status signal; The impedance compensation verification circuit is used to detect the matching degree between the actual output impedance and the target impedance.

9. The system according to claim 6, wherein: The FPGA of the multi-objective collaborative control device is configured as follows: Perform parallel computation of the MPC algorithm in the pre-synchronization phase; Achieve nanosecond-level updates of weight factors in the hybrid control phase; The fast switching of positive / negative sequence impedance is completed in the dynamic damping stage.

Images