A multi-time scale power oscillation suppression system and method for network-constructed energy storage

Abstract

This invention provides a multi-timescale power oscillation suppression system and method for grid-based energy storage. The system includes: a signal acquisition and processing module for outputting real-time grid angular frequency signals, dq-axis voltage components, dq-axis current components, and instantaneous power signals; a multi-timescale stabilizer including a power control timescale stabilizer, a voltage control timescale stabilizer, and a current control timescale stabilizer arranged in parallel, which independently extract oscillation characteristics and generate corresponding compensation signals for the low-frequency electromechanical oscillation band, the medium-frequency controller interaction band, and the high-frequency electromagnetic oscillation band, respectively; a battery management system for real-time acquisition of battery state of charge, temperature, and power boundary information, and transmission to the multi-timescale stabilizer to dynamically adjust the amplitude limiting boundary of the compensation signal; and a grid-based main control unit for injecting the compensation signal into the corresponding control loop to generate a modulation signal based on a virtual synchronous generator control strategy.

Description

Technical Field

[0001] This invention belongs to the field of new energy grid connection and energy storage control technology, specifically relating to a multi-timescale power oscillation suppression system and method for grid-type energy storage. Background Technology

[0002] Recently, the proportion of new energy power generation, represented by wind power and photovoltaics, in the power system has been increasing. Due to the gradual replacement of traditional synchronous generators, the power grid exhibits characteristics of low inertia and weak damping, posing a severe challenge to the stable operation of the power system. To support the frequency and voltage stability of the new power system, grid-forming energy storage technology with active support capabilities has emerged. Grid-forming energy storage typically employs a virtual synchronous generator (VSG) control strategy, simulating the mechanical and electrical characteristics of a synchronous generator to provide necessary inertia and damping support for the grid, and has become a research hotspot in the field of new energy grid connection. However, grid-forming energy storage systems are typical high-order nonlinear systems, with control loops covering multiple time scales such as power loop, voltage loop, and current loop. Under weak grid conditions or in scenarios with multiple generators operating in parallel, due to changes in line impedance and strong coupling between control parameters, the system is highly susceptible to power oscillations at multiple time scales. For example, the dominant pole of the power loop may induce low-frequency electromechanical oscillations (0.1-2Hz), while the interaction between the voltage and current inner loops and the grid impedance may induce medium- and high-frequency electromagnetic oscillations. This wide-frequency oscillation not only limits the output capacity of the energy storage system, but in severe cases it can even lead to system instability and grid disconnection, posing a significant threat to grid security.

[0003] Existing oscillation suppression methods typically employ fixed virtual inertia or additional damping control. In fixed virtual inertia control strategies, the system simulates the rotating inertia response of a synchronous generator by setting a constant inertia coefficient to slow down the rate of frequency change; while additional damping control typically introduces differential elements or damping coefficients into the power loop or voltage loop to suppress power fluctuations by dissipating oscillation energy.

[0004] However, existing technologies have two significant limitations: First, traditional single-parameter tuning struggles to simultaneously consider the dynamic performance across different frequency bands. For example, while increasing damping can suppress low-frequency oscillations, it may weaken the ability to suppress high-frequency interference, lacking a frequency-band adaptive adjustment mechanism. Second, most existing control strategies are based on the assumption of an ideal DC voltage source, assuming that the DC side has unlimited power throughput, ignoring the nonlinear physical characteristics of the front-end battery energy storage unit constrained by state of charge (SOC), temperature, and power boundaries. When the control algorithm outputs high-power commands to smooth oscillations, a lack of decoupling processing for different frequency band characteristics and awareness of battery physical boundaries can easily lead to oscillation suppression failure, DC bus voltage exceeding limits, or even triggering system protection tripping. Summary of the Invention

[0005] This invention proposes a multi-timescale power oscillation suppression system and method for grid-type energy storage, which solves the problem that existing technologies are prone to oscillation suppression failure or system protection tripping under complex operating conditions.

[0006] To address the aforementioned technical problems, this invention provides a multi-timescale power oscillation suppression system for grid-based energy storage, comprising:

[0007] The signal acquisition and processing module is used to acquire and process the voltage and current signals of the power grid, and output the real-time angular frequency signal, dq-axis voltage component and instantaneous active power signal of the power grid.

[0008] The multi-time-scale stabilizer includes a power-controlled time-scale stabilizer, a voltage-controlled time-scale stabilizer, and a current-controlled time-scale stabilizer arranged in parallel.

[0009] The power control time-scale stabilizer extracts oscillation characteristics and generates corresponding compensation signals based on the real-time angular frequency signal of the power grid for the low-frequency electromechanical oscillation band.

[0010] The voltage-controlled time-scale stabilizer extracts oscillation characteristics and generates corresponding compensation signals based on the instantaneous active power signal for the intermediate frequency controller's interactive frequency band.

[0011] The current-controlled time-scale stabilizer is based on the dq-axis voltage component, extracts oscillation characteristics for the high-frequency electromagnetic oscillation band and generates corresponding compensation signals;

[0012] The battery management system is used to collect the battery's state of charge, temperature, and power boundary information in real time, and transmit it to the multi-timescale stabilizer to dynamically adjust the amplitude limiting boundary of the compensation signal.

[0013] The network-type main control unit is used to inject the compensation signal into the corresponding control loop to generate a modulation signal based on the virtual synchronous generator control strategy.

[0014] Preferably, the current-controlled time-scale stabilizer adopts a dual-channel cross-decoupling structure with a 2×2 virtual admittance matrix, comprising:

[0015] The d-axis processing branch includes a d-axis high-pass filter, a d-axis self-admittance module, and a dq-axis mutual admittance module.

[0016] The q-axis processing branch includes a q-axis high-pass filter, a q-axis self-admittance module, and a qd-axis mutual admittance module.

[0017] The output of the d-axis self-admittance module is superimposed with the output of the qd-axis mutual admittance module to form a direct-axis current compensation signal, and the output of the q-axis self-admittance module is superimposed with the output of the dq-axis mutual admittance module to form a quadrature-axis current compensation signal. The equivalent output impedance of the system in the high-frequency band is reshaped by the cross-control of the virtual admittance matrix.

[0018] Preferably, the power control time-scale stabilizer includes a normalization stage, a low-pass filter stage, a DC blocking stage, a phase compensation stage, a gain amplification stage, and a dynamic limiting stage, which are cascaded in sequence.

[0019] The control input terminal of the dynamic limiting circuit receives the state of charge signal, temperature signal and power limit signal from the battery management system. It dynamically adjusts the upper and lower boundaries of the output limiting according to the current charging and discharging capacity of the battery. When the battery state of charge or temperature approaches the limit threshold, the limiting boundary is automatically contracted and an inertia compensation signal is output.

[0020] Preferably, the voltage-controlled time-scale stabilizer includes a bandpass filter, a DC blocking stage, a first phase compensation stage, a second phase compensation stage, a gain amplification stage, and an amplitude limiting stage, which are cascaded in sequence.

[0021] The bandpass filter is used to extract the power oscillation component within the target frequency band from the instantaneous active power signal;

[0022] The first phase compensation stage and the second phase compensation stage use a lead-lag correction transfer function to perform two-stage phase shaping, compensate for the phase lag introduced by the detection and control loop, and output a quadrature-axis voltage compensation signal.

[0023] Preferably, the signal acquisition and processing module includes: a phase-locked loop unit, a coordinate transformation unit, and a power calculation unit;

[0024] The phase-locked loop unit is used to track the phase information of the grid voltage at the point of common coupling and output the real-time angular frequency signal of the grid.

[0025] The coordinate transformation unit uses Park transformation to convert the voltage signal in the three-phase stationary coordinate system into the dq-axis voltage component in the synchronous rotating coordinate system.

[0026] The power calculation unit calculates the instantaneous active power signal based on instantaneous power theory.

[0027] Preferably, the network-type main control unit includes:

[0028] Active power-frequency control loop: used to receive active power setpoint, instantaneous active power signal and inertia compensation signal, and output virtual synchronous generator angular frequency signal;

[0029] Reactive power-voltage control loop: The d-axis voltage reference value is calculated through a droop control algorithm;

[0030] d-axis voltage loop PI controller and d-axis current loop PI controller: used to output a d-axis modulated voltage signal after superimposing the direct-axis current compensation signal in the d-axis channel;

[0031] q-axis voltage loop PI controller and q-axis current loop PI controller: used to output a q-axis modulated voltage signal after superimposing the quadrature axis voltage compensation signal and the quadrature axis current compensation signal in the q-axis channel;

[0032] Inverse coordinate transformation unit: used to restore the d-axis and q-axis modulated voltage signals to three-phase AC modulated signals.

[0033] Preferably, the multi-timescale stabilizer further includes a state constraint decision unit, which is used to receive state information from the battery management system, monitor the overall operating status of the system in real time, determine whether the system meets the grid connection conditions, and output start / stop commands to the battery management system.

[0034] Preferably, the battery management system includes an analog front-end sampling circuit, an A / D conversion circuit, a microcontroller unit, an electrical isolation circuit, a CAN transceiver, and a data storage unit;

[0035] The microcontroller unit is used to perform state of charge estimation, power boundary calculation and fault diagnosis procedures.

[0036] The CAN transceiver is connected to the microcontroller unit via the electrical isolation circuit, enabling bidirectional data transmission with the multi-timescale stabilizer.

[0037] Preferably, the expression for the lead-lag compensation transfer function is:

[0038] ;

[0039] In the formula, , It is a time constant; For the Laplace operator.

[0040] This invention also provides a method for suppressing multi-timescale power oscillations in grid-based energy storage, which is based on the aforementioned multi-timescale power oscillation suppression system for grid-based energy storage and includes the following steps:

[0041] Step S1: Collect and process the voltage and current signals of the power grid, output the real-time angular frequency signal of the power grid, the dq axis voltage component and the instantaneous active power signal, and collect the state of charge, temperature and power boundary information of the battery.

[0042] Step S2: For the low-frequency electromechanical oscillation band, generate an inertia compensation signal from the real-time angular frequency signal of the power grid; for the medium-frequency controller interaction band, generate a quadrature-axis voltage compensation signal from the instantaneous active power signal; for the high-frequency electromagnetic oscillation band, use virtual admittance matrix cross-decoupling to generate a dq-axis current compensation signal from the dq-axis voltage component.

[0043] Step S3: Dynamically adjust the amplitude limiting boundary of the compensation signal based on the battery status information;

[0044] Step S4: Inject each compensation signal into the corresponding link of the virtual synchronous generator control loop to generate a modulation signal to drive the energy storage converter.

[0045] The advantages of this invention include at least the following:

[0046] 1. By setting independent stabilizers for three time scales—power, voltage, and current—in parallel, feature extraction and compensation are performed for low-frequency electromechanical oscillations, the intermediate frequency controller interaction band, and high-frequency electromagnetic oscillations, respectively, overcoming the limitation of traditional single-parameter tuning that cannot cover the entire frequency band.

[0047] 2. The introduction of a battery management system provides real-time feedback on SOC, temperature, and power boundary information, and dynamically adjusts the compensation signal limit to avoid the risk of DC bus voltage exceeding the limit or protection tripping under the ideal DC source assumption.

[0048] 3. The compensation signal is directly injected into the VSG control loop, which achieves the organic unity of oscillation suppression and basic control functions while maintaining the active support capability of grid-type energy storage. Attached Figure Description

[0049] Figure 1 This is a topology diagram of a grid-type energy storage system integrated into an AC power grid according to an embodiment of the present invention;

[0050] Figure 2 This is an internal control structure diagram of the signal acquisition and processing module in an embodiment of the present invention;

[0051] Figure 3 This is a diagram of the internal control structure of the network-type main control unit in an embodiment of the present invention;

[0052] Figure 4 This is a diagram of the internal control architecture of the multi-timescale stabilizer in an embodiment of the present invention;

[0053] Figure 5 This is a diagram of the internal control logic structure of the power control time-scale stabilizer in this embodiment of the invention;

[0054] Figure 6 This is a diagram of the internal control logic structure of the voltage-controlled time-scale stabilizer in this embodiment of the invention.

[0055] Figure 7 This is a diagram of the internal control logic structure of the current-controlled time-scale stabilizer in an embodiment of the present invention;

[0056] Figure 8 This is a hardware architecture and control logic diagram of the battery management system (BMS) in an embodiment of the present invention;

[0057] Figure 9 This is a comparison diagram of active power before and after adding the oscillation suppression system in an embodiment of the present invention;

[0058] Figure 10 This is a comparison diagram of reactive power before and after adding the oscillation suppression system in an embodiment of the present invention;

[0059] Figure 11 This is a frequency comparison diagram before and after adding the oscillation suppression system in an embodiment of the present invention;

[0060] Figure 12 This is a comparison chart of the oscillation divergence and suppression effects of the system in this embodiment of the invention under critical instability parameters. Detailed Implementation

[0061] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the protection scope of the present invention.

[0062] like Figure 1 As shown, this embodiment of the invention provides a multi-timescale power oscillation suppression system for grid-based energy storage. The system comprises a DC battery module 1, a DC bus capacitor 2, an energy storage converter (PCS) 3, a filter inductor 4, a filter capacitor 5, a grid-side inductor 6, a public power grid 7, a signal acquisition and processing module 8, a grid-based main control unit 9, a drive circuit 10, a multi-timescale stabilizer 11, and a battery management system (BMS) 12.

[0063] like Figure 2 The diagram shown is an internal control structure diagram of the signal acquisition and processing module according to an embodiment of the present invention. The signal acquisition and processing module 8 includes a phase-locked loop unit 81, a coordinate transformation unit 82, and a power calculation unit 83.

[0064] The phase-locked loop unit 81 is used to acquire the three-phase voltage signal of the point of common coupling in real time, and to track the phase information of the grid voltage in real time through internal closed-loop control, and output the real-time angular frequency of the grid 810.

[0065] The coordinate transformation unit 82 receives three-phase voltage signals, three-phase current signals, and a virtual synchronous generator angular frequency signal 910 from the grid-type main control unit 9. After integrating to obtain the synchronous phase, it uses the Park transformation principle to convert the AC quantities in the three-phase stationary coordinate system into the direct-axis voltage component 820 in the synchronous rotating coordinate system. ), cross-axis voltage component 821 ( ), Direct-axis current component 822 ( ) and quadrature axis current component 823 ( ).

[0066] The power calculation unit 83 receives the direct-axis voltage component 820, the quadrature-axis voltage component 821, the direct-axis current component 822, and the quadrature-axis current component 823, and calculates the current instantaneous active power signal 830 and instantaneous reactive power signal 831 of the system based on instantaneous power theory.

[0067] like Figure 3 The diagram shows the internal control structure of the network-type main control unit. The network-type main control unit 9 mainly includes an active power-frequency control loop 91, a reactive power-voltage control loop 92, a d-axis voltage loop PI controller 93, a d-axis current loop PI controller 94, a q-axis voltage loop PI controller 95, and a q-axis current loop PI controller 96.

[0068] Active power-frequency control loop 91 receives active power setpoint of the system The system receives the instantaneous active power signal 830 and the inertia compensation signal 1360 from the multi-timescale stabilizer 11; the virtual synchronous generator angular frequency signal 910 is calculated by the virtual synchronous generator (VSG) active power control algorithm.

[0069] The reactive power-voltage control loop 92 receives the reactive power setpoint from the system. The instantaneous reactive power signal 831 is fed back; the d-axis voltage reference value 920 is calculated through the droop control algorithm, and the q-axis voltage reference value is usually set to 0.

[0070] In the d-axis control channel, the difference between the d-axis voltage reference value 920 and the feedback direct-axis voltage component 820 is input to the d-axis voltage loop PI controller 93, which outputs the d-axis current reference value 930. The difference between this reference value and the feedback direct-axis current component 822 is superimposed with the direct-axis current compensation signal 1540 from the multi-time-scale stabilizer 11, and then input to the d-axis current loop PI controller 94 to calculate the d-axis modulated voltage signal 940.

[0071] In the q-axis control channel, the difference between the q-axis voltage reference value and the feedback quadrature-axis voltage component 821 is calculated, and then superimposed with the quadrature-axis voltage compensation signal 1460 from the multi-time-scale stabilizer 11. This result is input to the q-axis voltage loop PI controller 95, which outputs the q-axis current reference value 950. In this embodiment of the invention, the q-axis voltage reference value is set to 0. The difference between this reference value and the feedback quadrature-axis current component 823 is calculated, and then superimposed with the quadrature-axis current compensation signal 1580 from the multi-time-scale stabilizer 11. This result is input to the q-axis current loop PI controller 96, which calculates the q-axis modulated voltage signal 960.

[0072] The inverse coordinate transformation unit 97 receives the d-axis modulated voltage component 940, the q-axis modulated voltage component 960, and the virtual synchronous generator angular frequency signal 910. Using the inverse Park transformation principle, it restores the DC voltage component in the rotating coordinate system to the three-phase AC modulated signal 970 in the three-phase stationary coordinate system and sends it to the SVPWM drive module 10 to generate the switching pulse 100.

[0073] like Figure 4 The diagram shows the internal control architecture of the multi-time-scale stabilizer. The multi-time-scale stabilizer 11 includes a power control time-scale stabilizer 13, a voltage control time-scale stabilizer 14, a current control time-scale stabilizer 15, and a state constraint decision unit 16, all configured in parallel.

[0074] The power control time-scale stabilizer 13 receives the real-time angular frequency signal 810 of the power grid from the signal acquisition and processing module 8, as well as the temperature signal, state of charge (SOC) signal, and power limit signal from the battery management system 12. It dynamically adjusts the control boundary according to the battery state and outputs the inertia compensation signal 1360 after internal calculation and processing.

[0075] The voltage-controlled time-scale stabilizer 14 receives the instantaneous active power signal 830 from the signal acquisition and processing module 8, and outputs the quadrature-axis voltage compensation signal 1460 after internal processing.

[0076] The current-controlled time-scale stabilizer 15 receives the direct-axis voltage component 820 and the quadrature-axis voltage component 821 from the signal acquisition and processing module 8. After internal cross-decoupling processing, it outputs the direct-axis current compensation signal 1540 and the quadrature-axis current compensation signal 1580. The state constraint decision unit 16 receives state information from the battery management system 12 and is responsible for monitoring the overall operating status of the system in real time, determining whether the system meets the grid-connected operation conditions, and outputting start / stop commands 160 by the logic decision unit.

[0077] like Figure 5The diagram shows the internal control logic structure of the power control time-scale stabilizer. The power control time-scale stabilizer 13 is mainly composed of a normalization stage 131, a low-pass filter stage 132, a DC blocking stage 133, a phase compensation stage 134, a gain amplification stage 135, and a dynamic limiting stage 136 cascaded together.

[0078] Normalization step 131 receives the real-time angular frequency signal 810 of the power grid from the signal acquisition and processing module 8, and divides it by The signal is converted into a frequency signal. Then, it passes through a low-pass filter 132 and a DC blocking filter 133 to process the frequency signal, filter out high-frequency noise and eliminate DC components, and extract the frequency change that reflects the low-frequency oscillation characteristics of the power grid.

[0079] Phase compensation stage 134 employs a lead-lag compensation transfer function to perform phase lead compensation on the frequency change, thereby compensating for the phase lag introduced by the system's measurement and control loops. The expression for the lead-lag compensation transfer function is as follows:

[0080] ;

[0081] In the formula, , It is a time constant; For the Laplace operator.

[0082] After phase compensation, the signal enters the gain amplification stage 135 and is multiplied by a preset gain coefficient. The amplified signal enters the dynamic limiting stage 136. This stage receives temperature signals, state of charge (SOC) signals, and power limit signals from the battery management system 12 at its control input. Based on the battery's current charge / discharge capacity, it dynamically adjusts the upper and lower boundaries of the output limiting: when the battery's SOC or temperature approaches the limit threshold, the limiting boundaries automatically shrink. After processing by the dynamic limiting stage 136, the amplified signal outputs the final inertia compensation signal 1360.

[0083] like Figure 6 The diagram shows the internal control logic structure of the voltage-controlled time-scale stabilizer. The voltage-controlled time-scale stabilizer 14 is mainly composed of a cascaded bandpass filter 141, a DC blocking filter 142, a first phase compensation filter 143, a second phase compensation filter 144, a gain amplification filter 145, and an amplitude limiting filter 146.

[0084] The bandpass filter stage 141 receives the instantaneous active power signal 830 from the signal acquisition and processing module 8, and uses a bandpass filter (BPF) to initially extract the power oscillation components within the target frequency band. In this embodiment of the invention, the target frequency band is set to 2Hz-50Hz.

[0085] The DC blocking stage 142 receives the bandpass-filtered signal and further filters out DC bias and low-frequency drift using a high-pass transfer function, ensuring that the subsequent compensation signal is a pure AC oscillation component. The expression for the high-pass transfer function is:

[0086] ;

[0087] In the formula, is the time constant of the high-pass filter.

[0088] The first phase compensation stage 143 and the second phase compensation stage 144 are connected in series, both employing a lead-lag compensation transfer function. The expressions for the lead-lag compensation transfer functions of the first phase compensation stage 143 and the second phase compensation stage 144 are as follows:

[0089] ;

[0090] ;

[0091] In the formula, , The time constant of the first phase compensation stage; , This is the time constant of the second phase compensation stage.

[0092] The signal output from the DC blocking stage 142 undergoes two stages of phase shaping, namely the first phase compensation stage 143 and the second phase compensation stage 144, to provide a sufficient phase lead angle and fully compensate for the phase lag introduced by the detection and control loops. The signal after the two-stage phase compensation enters the gain amplification stage 145 and is multiplied by a preset proportional gain coefficient. The amplified signal is input to amplitude limiting circuit 146, which has fixed upper limits for positive and negative voltages. After limiting, the final output is a quadrature-axis voltage compensation signal 1460.

[0093] like Figure 7 The diagram shows the internal control logic structure of the current-controlled time-scale stabilizer. The current-controlled time-scale stabilizer 15 adopts a dual-channel cross-decoupling structure, mainly including a d-axis high-pass filter 151, a d-axis self-admittance module 152, a dq-axis mutual admittance module 153, a d-axis current limiting module 154, a q-axis high-pass filter 155, a qd-axis mutual admittance module 156, a q-axis self-admittance module 157, and a q-axis current limiting module 158.

[0094] In the d-axis signal processing branch, the d-axis high-pass filter 151 receives the direct-axis voltage component 820, filters out the DC component, and extracts the direct-axis high-frequency voltage disturbance. This disturbance is split into two paths: one path is input to the d-axis self-admittance module 152 for calculation, and the other path is input to the dq-axis mutual admittance module 153 for cross-coupling calculation.

[0095] In the q-axis signal processing branch, the q-axis high-pass filter 155 receives the quadrature-axis voltage component 821, filters out the DC component, and extracts the quadrature-axis high-frequency voltage disturbance. This disturbance is also split into two paths: one path is input to the q-axis self-admittance module 157 for calculation, and the other path is input to the qd-axis mutual admittance module 156 for cross-coupling calculation.

[0096] During the cross-decoupling output stage, the output signal 1520 of the d-axis self-admittance module 152 and the output signal 1560 of the qd-axis mutual admittance module 156 are superimposed. After being limited by the d-axis current limiting circuit 154, a direct-axis current compensation signal 1540 is generated. Simultaneously, the output signal 1570 of the q-axis self-admittance module 157 and the output signal 1530 of the dq-axis mutual admittance module 153 are superimposed. After being limited by the q-axis current limiting circuit 158, a quadrature-axis current compensation signal 1580 is generated. Through the cross-control of the aforementioned virtual admittance matrix, the equivalent output impedance of the system in the high-frequency band is reshaped, suppressing high-frequency oscillations.

[0097] like Figure 8 The diagram shows the hardware architecture and control logic of the battery management system (BMS). The battery management system 12 mainly consists of an analog front-end sampling circuit 121, an A / D conversion circuit 122, a microcontroller unit (MCU) 123, an electrical isolation circuit 124, a CAN transceiver 125, and a data storage unit 126.

[0098] The analog front-end sampling circuit 121 is directly connected to each individual cell of the DC battery module 1 to collect the battery pack's terminal voltage, loop current, and temperature signals at key points in real time. The collected analog signals are converted into digital signals by the A / D conversion circuit 122 through high-precision analog-to-digital conversion and input to the microcontroller unit 123.

[0099] The microcontroller unit 123 serves as the core processing unit, responsible for executing SOC estimation, power boundary calculation, and fault diagnosis procedures. For data storage, the microcontroller unit 123 is bidirectionally connected to the data storage unit 126, used for real-time storage of operating data and historical fault records.

[0100] In terms of external communication and control, the CAN transceiver 125 is connected to the microcontroller 123 via the electrical isolation circuit 124 to achieve bidirectional data transmission: on the one hand, it sends the temperature, state of charge (SOC), power limit and fault status calculated by the microcontroller 123 to the external multi-timescale stabilizer 11 as the constraint basis for the multi-timescale stabilizer 11 to perform power distribution and dynamic limiting; on the other hand, it receives the system start-stop command 160 issued from the network-type main control unit, and the microcontroller 123 controls the on and off of the internal high-voltage relay according to the command, thereby realizing the management of the charging and discharging circuit of the energy storage battery system.

[0101] Through the aforementioned hardware architecture and control logic design, this embodiment of the invention presents a grid-based energy storage multi-timescale power oscillation suppression system. In this system, the energy storage battery module and converter constitute the physical execution foundation; the battery management system senses the physical boundaries in real time and uploads constraint commands; the grid-based main control unit establishes the system's voltage and frequency references using a virtual synchronization control strategy, achieving self-synchronization grid connection; and the multi-timescale stabilizer serves as a parallel compensation link, addressing grid frequency, active power, and... The oscillation characteristics of different frequency bands are extracted from the shaft voltage, and after decoupling operations, a compensation signal is generated and injected into the main control loop. Ultimately, under the safety constraints of the battery management system, the system achieves full-link coordinated control from the physical layer to the control layer, and from low-frequency electromechanical dynamics to high-frequency electromagnetic transients, thus completing the active suppression of grid oscillations.

[0102] like Figure 9 As shown in the figure, the solid black line represents the active power response waveform without the oscillation suppression system, and the dashed blue line represents the active power response waveform after incorporating the oscillation suppression system of this embodiment. It is clearly visible in the figure that during system operation, especially at the step changes of 0.5s power command down and 1.0s power command up, the system without oscillation suppression exhibits significant underdamping characteristics. The active power waveform shows a large overshoot accompanied by a significant oscillation decay process, resulting in a longer time for the system to re-stabilize at the target power value, leading to poor dynamic performance. In contrast, after incorporating the oscillation suppression system, the equivalent damping of the system is significantly enhanced. Under the same power step disturbance, the active power waveform transitions smoothly, essentially eliminating overshoot and repeated oscillations. It can quickly converge and accurately stabilize at the given power command value, thus verifying that the power control time-scale stabilizer in this embodiment can effectively suppress low-frequency electromechanical oscillations and significantly improve the dynamic stability and speed of the grid-type energy storage system.

[0103] like Figure 10As shown in the figure, the solid black line represents the reactive power response waveform without the oscillation suppression system, and the dashed blue line represents the reactive power response waveform after incorporating the oscillation suppression system of this embodiment. It can be clearly observed from the figure that when the system encounters operating disturbances at 0.5s and 1.0s, the system without oscillation suppression exhibits severe dynamic fluctuations, with the reactive power amplitude oscillating significantly between 3000Var and 7000Var, showing significant overshoot and underdamping. This means that the system lacks sufficient damping to smooth out mid-frequency interactive oscillations. In contrast, after incorporating the oscillation suppression system, the system can quickly reshape the output impedance through virtual damping, significantly suppressing reactive power surges and oscillations. The waveform transition is smooth with minimal overshoot, and it can quickly and smoothly converge to a steady-state value of 5000Var. This result demonstrates the effectiveness of the voltage-controlled time-scale stabilizer in suppressing mid-frequency interactive oscillations in this embodiment, significantly improving the voltage support capability and reactive power regulation stability of the grid-type energy storage system under weak grid conditions.

[0104] like Figure 11 As shown in the figure, the solid black line represents the frequency response waveform without the oscillation suppression system, and the dashed blue line represents the frequency response waveform after incorporating the oscillation suppression system of this embodiment. It can be seen from the figure that when the system is in a dynamic adjustment process, such as during sudden changes in operating conditions at 0.5s and 1.0s, the system without oscillation suppression experiences drastic frequency fluctuations and a large frequency deviation, with a maximum spike to approximately 50.3Hz and a minimum drop to approximately 49.7Hz. It only slowly returns to the 50Hz reference value after multiple oscillations, indicating insufficient system inertia and weak disturbance rejection capability. In contrast, after incorporating the oscillation suppression system, thanks to the virtual inertia and damping support provided by the grid-based control strategy, the dynamic deviation of the system frequency is significantly compressed, and the waveform can smoothly and quickly recover to the 50Hz steady state. This verifies the significant effect of the system of this embodiment in improving grid frequency stability and enhancing system inertia support.

[0105] like Figure 12As shown in the figure, the solid black line represents the power response without the oscillation suppression system, indicating an unstable state, while the dashed blue line represents the power response with the oscillation suppression system of this embodiment, indicating a stable state. The figure clearly shows that when the system is under critical parameter conditions and encounters a small disturbance at 0.5s, the system without oscillation suppression, lacking sufficient damping support, exhibits typical divergent oscillation characteristics in its active power. The oscillation amplitude increases continuously over time, indicating that the system has entered a negatively damped unstable region, which, without intervention, will lead to grid disconnection. In contrast, after adding the oscillation suppression system, the multi-timescale stabilizer quickly intervenes and reshapes the system impedance characteristics, forcibly introducing positive damping. This causes the power oscillation, which was about to diverge, to rapidly decay and converge to a steady state within a few cycles. This powerfully demonstrates that the system of this embodiment not only optimizes dynamic performance but also possesses the core capability to prevent system oscillation instability and ensure safe equipment operation under extreme critical conditions.

[0106] This invention also provides a method for suppressing multi-timescale power oscillations in grid-based energy storage, based on the aforementioned multi-timescale power oscillation suppression system for grid-based energy storage, comprising the following steps:

[0107] Step S1: Collect the voltage and current signals of the power grid, obtain the real-time angular frequency signal, dq-axis voltage and current components and instantaneous power signal of the power grid, and collect the state of charge, temperature and power boundary information of the battery.

[0108] Step S2: For the low-frequency electromechanical oscillation band, generate an inertia compensation signal from the grid angular frequency signal; for the medium-frequency controller interaction band, generate a quadrature-axis voltage compensation signal from the instantaneous active power signal; for the high-frequency electromagnetic oscillation band, use virtual admittance matrix cross-decoupling to generate a dq-axis current compensation signal from the dq-axis voltage component.

[0109] Step S3: Dynamically adjust the limiting boundary of the compensation signal based on the battery status information.

[0110] Step S4: Inject each compensation signal into the corresponding link of the virtual synchronous generator control loop to generate a modulation signal to drive the energy storage converter.

[0111] By introducing the grid-type main control unit into the virtual synchronous generator control strategy, the power electronic converter is endowed with inertia and damping characteristics similar to traditional synchronous generators, enabling it to actively support grid voltage and frequency. This effectively solves the problem of grid inertia deficiency and insufficient strength caused by the high proportion of new energy access, and significantly improves the system's operational stability under weak grid and islanded conditions.

[0112] Based on this, the embodiments of the present invention construct a parallel multi-timescale stabilizer architecture, covering three key timescales: power control, voltage control, and current control. Through feature extraction and independent control for different frequency bands, it can simultaneously and accurately suppress low-frequency electromechanical oscillations, mid-frequency controller interactive oscillations, and high-frequency harmonic resonances, avoiding inter-band coupling interference that may be caused by single-dimensional control, and achieving comprehensive management of wide-frequency power system oscillations. Furthermore, it addresses the oscillations that are prone to occur in the high-frequency band. To address the shaft coupling instability problem, this invention introduces a current-controlled timescale. The virtual admittance matrix effectively reshapes the high-frequency output impedance characteristics of the converter by establishing a cross-decoupling path between the direct axis and the quadrature axis, thereby enhancing the system's robustness to high-frequency resonance and parameter perturbations of the power grid.

[0113] Furthermore, this embodiment of the invention establishes a two-way deep interaction mechanism between the battery management system and the grid-type controller. The state constraint decision unit can dynamically adjust the amplitude limiting threshold or start / stop state of the oscillation suppression strategy based on the battery's real-time state of charge, temperature, and power boundary. This mechanism maximizes the value of energy storage in stabilizing the grid while effectively preventing the risks of battery overcharging, over-discharging, or overheating caused by forcibly suppressing oscillations, thereby ensuring the efficient, safe, and reliable operation of the energy storage system throughout its entire life cycle.

[0114] The technical features of the above embodiments can be combined arbitrarily. For the sake of brevity, not all possible combinations of the technical features in the above embodiments are described; only preferred embodiments of the present invention are illustrated. The descriptions are relatively specific and detailed, but they should not be construed as limiting the scope of the present invention. As long as the combination of these technical features does not contradict each other, it should be considered within the scope of this specification.

[0115] It should be noted that those skilled in the art can make various modifications and improvements without departing from the inventive concept, and these all fall within the scope of protection of this invention. Therefore, the scope of protection of this invention should be determined by the appended claims.

Claims

1. A multi-time scale power oscillation suppression system for network-constructed energy storage, characterized in that, include: Signal acquisition and processing module, multi-timescale stabilizer, battery management system and grid-type main control unit; The signal acquisition and processing module is used to acquire and process the voltage and current signals of the power grid, and output the real-time angular frequency signal, dq-axis voltage component and instantaneous active power signal of the power grid. The multi-time-scale stabilizer includes a power-controlled time-scale stabilizer, a voltage-controlled time-scale stabilizer, and a current-controlled time-scale stabilizer arranged in parallel. The power control time-scale stabilizer extracts oscillation characteristics and generates corresponding compensation signals based on the real-time angular frequency signal of the power grid for the low-frequency electromechanical oscillation band. The voltage-controlled time-scale stabilizer extracts oscillation characteristics and generates corresponding compensation signals based on the instantaneous active power signal for the intermediate frequency controller's interactive frequency band. The current-controlled time-scale stabilizer is based on the dq-axis voltage component, extracts oscillation characteristics for the high-frequency electromagnetic oscillation band and generates corresponding compensation signals; The battery management system is used to collect the battery's state of charge, temperature, and power boundary information in real time, and transmit it to the power control time scale stabilizer to dynamically adjust the limiting boundary of the compensation signal. The network-type main control unit is used to inject the compensation signal into the corresponding control loop to generate a modulation signal based on the virtual synchronous generator control strategy. 2.The multi-time scale power oscillation suppression system of network-constructed energy storage according to claim 1, wherein: The current-controlled time-scale stabilizer employs a dual-channel cross-decoupling structure with a 2×2 virtual admittance matrix, including: The d-axis processing branch includes a d-axis high-pass filter, a d-axis self-admittance module, and a dq-axis mutual admittance module. The d-axis high-pass filter receives the direct-axis voltage component, filters out the DC component, and outputs the direct-axis high-frequency voltage disturbance. The direct-axis high-frequency voltage disturbance is split into two paths: one path is input to the d-axis self-admittance module, and the other path is input to the dq-axis mutual admittance module. The q-axis processing branch includes a q-axis high-pass filter, a q-axis self-admittance module, and a qd-axis mutual admittance module. The q-axis high-pass filter receives the quadrature-axis voltage component, filters out the DC component, and outputs the quadrature-axis high-frequency voltage disturbance. The quadrature-axis high-frequency voltage disturbance is split into two paths: one path is input to the q-axis self-admittance module, and the other path is input to the qd-axis mutual admittance module. The output of the d-axis self-admittance module is superimposed with the output of the qd-axis mutual admittance module to form a direct-axis current compensation signal, and the output of the q-axis self-admittance module is superimposed with the output of the dq-axis mutual admittance module to form a quadrature-axis current compensation signal. The equivalent output impedance of the system in the high-frequency band is reshaped by the cross-control of the virtual admittance matrix. 3.The multi-time scale power oscillation suppression system of network-constructed energy storage according to claim 1, wherein: The power control time-scale stabilizer includes a normalization stage, a low-pass filter stage, a DC blocking stage, a phase compensation stage, a gain amplification stage, and a dynamic limiting stage, which are cascaded in sequence. The control input terminal of the dynamic limiting circuit receives the state of charge signal, temperature signal and power limit signal from the battery management system. It dynamically adjusts the upper and lower boundaries of the output limiting according to the current charging and discharging capacity of the battery. When the battery state of charge or temperature approaches the limit threshold, the limiting boundary is automatically contracted and an inertia compensation signal is output. 4.The multi-time scale power oscillation suppression system of network-constructed energy storage according to claim 1, wherein: The voltage-controlled time-scale stabilizer includes a bandpass filter, a DC blocking stage, a first phase compensation stage, a second phase compensation stage, a gain amplification stage, and an amplitude limiting stage, which are cascaded in sequence. The bandpass filter is used to extract the power oscillation component within the target frequency band from the instantaneous active power signal; The first phase compensation stage and the second phase compensation stage use a lead-lag correction transfer function to perform two-stage phase shaping, compensate for the phase lag introduced by the detection and control loop, and output a quadrature-axis voltage compensation signal.

5. The multi-time scale power oscillation suppression system of network-constructed energy storage according to claim 1, characterized in that: The signal acquisition and processing module includes: a phase-locked loop unit, a coordinate transformation unit, and a power calculation unit; The phase-locked loop unit is used to track the phase information of the grid voltage at the point of common coupling and output the real-time angular frequency signal of the grid. The coordinate transformation unit uses Park transformation to convert the voltage signal in the three-phase stationary coordinate system into the dq-axis voltage component and dq-axis current component in the synchronous rotating coordinate system. The power calculation unit receives the dq-axis voltage component and dq-axis current component output by the coordinate transformation unit, and calculates the instantaneous active power signal based on instantaneous power theory.

6. The multi-time scale power oscillation suppression system of network-constructed energy storage according to claim 1, characterized in that: The network-type main control unit includes: Active power-frequency control loop: used to receive active power setpoint, instantaneous active power signal and inertia compensation signal, and output virtual synchronous generator angular frequency signal; Reactive power-voltage control loop: The d-axis voltage reference value is calculated through a droop control algorithm; d-axis voltage loop PI controller and d-axis current loop PI controller: used to output a d-axis modulated voltage signal after superimposing the direct-axis current compensation signal in the d-axis channel; q-axis voltage loop PI controller and q-axis current loop PI controller: used to output a q-axis modulated voltage signal after superimposing the quadrature axis voltage compensation signal and the quadrature axis current compensation signal in the q-axis channel; Inverse coordinate transformation unit: used to restore the d-axis and q-axis modulated voltage signals to three-phase AC modulated signals.

7. The multi-timescale power oscillation suppression system for grid-type energy storage according to claim 1, characterized in that: The multi-timescale stabilizer also includes a state constraint decision unit, which is used to receive state information from the battery management system, monitor the overall operating status of the system in real time, determine whether the system meets the grid connection conditions, and output start / stop commands to the battery management system. 8.The multi-time scale power oscillation suppression system of network-constructed energy storage according to claim 1, wherein: The battery management system includes an analog front-end sampling circuit, an A / D conversion circuit, a microcontroller unit, an electrical isolation circuit, a CAN transceiver, and a data storage unit. The analog front-end sampling circuit is connected to the DC battery module to collect the battery pack's terminal voltage, loop current, and temperature signals in real time. The A / D conversion circuit receives the analog signal output by the analog front-end sampling circuit, performs analog-to-digital conversion, and then inputs it to the microcontroller unit. The microcontroller unit is used to perform state of charge estimation, power boundary calculation and fault diagnosis procedures. The microcontroller unit is bidirectionally connected to the data storage unit and is used to store operating data and historical fault records in real time. The CAN transceiver is connected to the microcontroller unit via the electrical isolation circuit, enabling bidirectional data transmission with the multi-timescale stabilizer. 9.The multi-time scale power oscillation suppression system of network-constructed energy storage according to claim 4, wherein: The expression for the lead-lag compensation transfer function is: ； wherein , is a time constant; is the Laplacian operator.

10. A method for suppressing multi-time scale power oscillation of grid-forming energy storage, implemented based on the multi-time scale power oscillation suppression system of any one of claims 1 to 9. Includes the following steps: Step S1: Collect and process the voltage and current signals of the power grid, output the real-time angular frequency signal of the power grid, the dq axis voltage component and the instantaneous active power signal, and collect the state of charge, temperature and power boundary information of the battery. Step S2: For the low-frequency electromechanical oscillation band, generate an inertia compensation signal from the real-time angular frequency signal of the power grid; for the medium-frequency controller interaction band, generate a quadrature-axis voltage compensation signal from the instantaneous active power signal; for the high-frequency electromagnetic oscillation band, use virtual admittance matrix cross-decoupling to generate a dq-axis current compensation signal from the dq-axis voltage component. Step S3: Dynamically adjust the amplitude limiting boundary of the compensation signal based on the battery status information; Step S4: Inject each compensation signal into the corresponding link of the virtual synchronous generator control loop to generate a modulation signal to drive the energy storage converter.

Images