Data-driven direct current capacitor group network self-synchronization new energy station control method

Abstract

The application discloses a kind of based on data-driven direct current capacitor group network self-synchronization new energy station control method.Method includes: constructing the direct current capacitor optimal self-synchronization unit of new energy station including data-driven DeePC controller, after direct current capacitor optimal self-synchronization, output modulated wave voltage phase signal;After voltage and current double closed loop control, obtain modulated wave voltage amplitude signal, and then obtain modulated wave voltage signal, after transformation modulation, act on the power switch device of converter of new energy station, realize new energy station control.The method of the application makes new energy station have optimal self-synchronization and networking function.At the same time, the stability of new energy station under weak power grid and the support ability to power grid are improved, the defects of traditional model-based network following control strategy are overcome, which is conducive to the safe and stable operation of new energy station.

Description

Technical Field

[0001] This invention relates to a control method for new energy power stations, specifically a data-driven control method for self-synchronizing DC capacitor banks in new energy power stations. Background Technology

[0002] Currently, new energy power plants, represented by wind and solar power, have become a core component in building new power systems and achieving a green and environmentally friendly energy transition. Traditional new energy power plants generally adopt a grid-following control strategy, relying on phase-locked loops (PLLs) to track the phase and frequency of the grid voltage in real time, thereby achieving synchronous operation with the grid. However, with the continuous increase in the penetration rate of new energy sources and the weakening of the grid strength, the system inertia and damping characteristics are deteriorating. This makes new energy power plants using grid-following control strategies prone to stability problems such as loss of synchronization and oscillations under grid disturbances, severely restricting the safe and stable operation of new power systems. Existing technologies are insufficient to achieve stable operation of new energy power plant clusters under weak grid conditions.

[0003] In weak grid environments, the phase-locked loop (PLL)-based synchronization control strategy for renewable energy power plants faces significant stability bottlenecks. In systems with low short-circuit ratios (SCRs), the PLL interacts strongly with the grid impedance, easily inducing broadband oscillations and leading to instability at the power plant level. Simultaneously, parameter perturbations, structural uncertainties, and model mismatches in system modeling cause significant performance degradation in traditional model-based control methods. To address the shortcomings of renewable energy power plants employing grid-following control strategies in weak grid environments, such as instability, poor grid support, and model mismatch, a novel control paradigm that proactively enhances system stability and is independent of PLLs is urgently needed. Summary of the Invention

[0004] To address the problems existing in the background technology, this invention provides a data-driven DC capacitor grid self-synchronization control method for renewable energy power plants. This method overcomes the shortcomings of traditional grid-following control methods, such as instability, poor adaptability, and weak support capabilities under weak grid conditions. Its core lies in using a data-driven optimized control method to autonomously generate synchronization signals based on real-time system operating data, significantly improving the operational stability and adaptability of renewable energy power plants such as photovoltaic and direct-drive wind turbines in weak grid environments. This method does not rely on phase-locked loops (PLLs), fundamentally eliminating the oscillation risk caused by traditional PLLs in weak grids. Optimal self-synchronization is achieved through DC capacitor voltage, effectively overcoming control deviations caused by model mismatch and enhancing the robustness of renewable energy power plants under parameter uncertainty conditions. Simultaneously, this method endows renewable energy power plants with grid characteristics, enabling rapid active power support when grid frequency drops occur, improving the dynamic characteristics of weak grids. This invention provides a reliable solution for the stable and efficient operation of renewable energy power plants in grids with low short-circuit ratios.

[0005] The technical solution adopted in this invention is:

[0006] The present invention provides a data-driven DC capacitor bank self-synchronizing renewable energy power station control method, comprising:

[0007] Step 1) Construct a DC capacitor optimal self-synchronization unit for the new energy power station, including a data-driven DeePC controller. Input the squared difference between the DC capacitor voltage of the new energy power station and the reference value, as well as the q-axis component of the output voltage, into the DC capacitor optimal self-synchronization unit. After processing, output the modulated wave voltage phase signal.

[0008] Step 2) Measure the output current, output voltage, and grid-side current of the renewable energy power station. The axial component is controlled by a dual closed-loop system of voltage and current, which includes both inner-loop current control and inner-loop current control, to obtain the modulated wave voltage amplitude signal. Axial components.

[0009] Step 3) Based on the modulation wave voltage phase signal and the modulation wave voltage amplitude signal... The shaft component obtains the modulated wave voltage signal, which is then transformed and modulated before being applied to the power switching devices of the converter in the new energy power station to realize the control of the new energy power station.

[0010] In step 1), the data-driven DeePC controller is specifically as follows:

[0011] ;

[0012] in, For decision variables; and Input and output slack variables are used respectively to enhance system robustness; and These are the inputs and outputs to be optimized for a stand-alone grid-connected system; and These are the input and output constraints for a stand-alone grid-connected system, respectively. For vectors quadratic form , =u or yr, P is the cost matrix, P=R or Q, R and Q are positive definite matrices and positive semi-definite matrices respectively, that is, the cost matrices of system input and system output; The second norm of vector S The square of S = or , = ; This serves as the reference output trajectory for a stand-alone grid-connected system. , and They are respectively , and The scaling factor, For coefficient Scaling regularization terms; It is a column vector; and These represent the first input and first output trajectories of a single-unit grid-connected system for predicting historical time periods. and These are the second input and second output trajectories of a single-unit grid-connected system for a predicted historical time period, with a duration of [duration missing]. ; and Each is the previous time of the current moment Input and output trajectories of a stand-alone grid-connected system collected over a period of time.

[0013] The input to a stand-alone grid-connected system includes the actual value of the DC capacitor voltage. Compared with reference value difference of squares and the q-axis component of the output voltage of the new energy power station The output includes the deviation of the control frequency signal. The single-unit grid-connected system is obtained by aggregating the equivalent value of new energy units, and the output of the DeePC controller is used as the input of the system.

[0014] The specific inputs and outputs to be optimized for the single-machine grid-connected system at time t are as follows:

[0015] ;

[0016] ;

[0017] ;

[0018] in, and The DeePC controller is a data-driven controller. Input and output at any moment and The DeePC controller is a data-driven controller. The first and second inputs at time 1; for Constantly control the deviation of the frequency signal; for The squared difference between the actual value and the reference value of the DC capacitor voltage at any given time. and The DC capacitor voltage is respectively at The actual value and reference value at that moment For the output voltage of the new energy power station Moment Axial components.

[0019] The data-driven DeePC controller employs a rolling optimization strategy, using rolling solutions to obtain... Input of a stand-alone grid-connected system that is optimized in real time That is, the deviation of the control frequency signal. , To control the time domain, the signal is then input into a stand-alone grid-connected system to obtain the control frequency signal. , , The current time is the rated frequency of the power grid. Update to the next sampling time The re-acquisition timing length is The latest system input and output trajectories are used to update the column vector. The optimization problem is then solved again to obtain the control frequency. In other words, this rolling optimization strategy achieves optimal self-synchronization of the DC capacitors in the renewable energy power plant.

[0020] In step 1), the DC capacitor optimal self-synchronization unit drives the control frequency signal output by the DeePC controller based on the data. The modulation wave voltage phase signal is obtained by performing integration. , .

[0021] In step 2), the output voltage and its reference value of the new energy power station, as well as the grid-side current, are... The axial component is controlled by the outer voltage loop to obtain current limiting, thereby obtaining the output current. The d-axis component reference value, together with the output current of the new energy power station, is used to obtain the d-axis component U of the modulated wave voltage amplitude signal after passing through the inner current loop control. td and q-axis component U tq .

[0022] In step 3), based on the modulated wave voltage phase signal The d-axis component U of the modulated wave voltage amplitude signal td and q-axis component U tq After performing the inverse Park transform, the modulated wave voltage signal U is obtained. tabc Then, a pulse width modulation signal is generated by a space vector pulse width modulation (SVPWM) generator and applied to the power switching devices of the converter in the new energy power station.

[0023] The data-driven DC capacitor grid self-synchronizing renewable energy power station control system of the present invention includes:

[0024] The unit processing module is used to construct the DC capacitor optimal self-synchronization unit of the new energy power station, which includes a data-driven DeePC controller. It inputs the square difference between the DC capacitor voltage of the new energy power station and the reference value, as well as the q-axis component of the output voltage, into the DC capacitor optimal self-synchronization unit, processes them, and outputs the modulated wave voltage phase signal.

[0025] The dual closed-loop control module is used to control the output current, output voltage, and grid-side current of the renewable energy power station. The axial component is controlled by a dual closed-loop system of voltage and current, which includes both inner-loop current control and inner-loop current control, to obtain the modulated wave voltage amplitude signal. Axial components.

[0026] The station control module is used to determine the modulation wave voltage phase signal and the modulation wave voltage amplitude signal. The shaft component obtains the modulated wave voltage signal, which is then transformed and modulated before being applied to the power switching devices of the converter in the new energy power station to realize the control of the new energy power station.

[0027] The electronic device of the present invention includes: a memory and a processor coupled to each other, wherein the memory stores program data, and the processor invokes the program data to execute the method described above.

[0028] The present invention provides a computer-readable storage medium having program data stored thereon, which, when executed by a processor, implements the method described above.

[0029] The data-enabled predictive control (DeePC) of this invention solves for the optimal control sequence online directly using system operating data, eliminating the reliance on precise mathematical models and demonstrating superior adaptability to dynamic system changes and unknown disturbances. The data-driven optimal self-synchronizing unit fundamentally avoids the stability problems of PLLs, making it suitable for renewable energy power plants such as photovoltaic or direct-drive wind turbines. Simultaneously, through optimized control of the DC capacitor voltage, the power plant as a whole becomes a network-type unit with autonomous adjustment capabilities, significantly enhancing the robustness and dynamic performance of the renewable energy power plant's interaction with the grid.

[0030] This invention addresses the problems of system instability, insufficient adaptability, and inadequate active support capabilities caused by phase-locked loops (PLLs) in traditional grid-connected control systems operating under weak grid conditions. The method employs a data-driven predictive control architecture, utilizing real-time system data to generate optimal control sequences online. This completely eliminates reliance on precise mathematical models, effectively preventing control performance degradation caused by model mismatch. Simultaneously, it achieves optimal self-synchronization of renewable energy power plants through a DC capacitor voltage self-synchronization mechanism, significantly improving the operational stability and dynamic response capabilities of photovoltaic, direct-drive wind turbine, and other renewable energy power plants in grids with low short-circuit ratios. Furthermore, this invention enables renewable energy power plants to exhibit excellent grid-connected operation characteristics, rapidly providing effective active power support during grid frequency disturbances, thereby enhancing the overall anti-interference capability and recovery stability of weak grids. This provides crucial technical support for the safe, stable, and efficient operation of new power systems.

[0031] The beneficial effects of this invention are:

[0032] This invention, through the data-driven predictive control (DeePC) framework, eliminates the reliance on phase-locked loops (PLLs) in grid-following control, fundamentally avoiding the instability problems caused by impedance coupling in weak grids. This significantly improves the adaptability and operational stability of renewable energy plants such as photovoltaic and direct-drive wind turbines under low short-circuit ratio grid conditions. Through the self-synchronization mechanism of DC capacitor voltage, optimal synchronization of renewable energy plants is achieved, giving them excellent dynamic response accuracy and disturbance rejection capabilities. Simultaneously, this invention endows renewable energy plants with grid control capabilities, enabling them to proactively provide active power support when grid frequency drops occur, enhancing system disturbance rejection performance and recovery capabilities. This improves the stability of renewable energy plants and their support for the grid under weak grid conditions, effectively supporting the safe and stable operation of weak grids. It promotes the key shift of renewable energy generation from "passive following" to "active grid connection," providing core technical support for the safe and stable operation of new power systems. Attached Figure Description

[0033] Figure 1 This is a flowchart of the method of the present invention;

[0034] Figure 2 This is a control block diagram of the method of the present invention;

[0035] Figure 3 This is a schematic diagram of a simulation model of a photovoltaic power station grid-connected system according to an embodiment of the present invention;

[0036] Figure 4 The simulation verification of the photovoltaic power station's output power step during the present invention includes the DC capacitor voltage and output active power waveforms. Figure 4 (a) is a simulation verification of the DC capacitor voltage waveform when the output power of the photovoltaic power station jumps. Figure 4 (b) is the active power waveform diagram when the output power of the photovoltaic power station jumps during simulation verification;

[0037] Figure 5 The simulation verification diagrams for the DC capacitor voltage and output active power waveforms of the photovoltaic power station during grid frequency drops are shown in this embodiment of the invention. Figure 5 (a) is a simulation verification diagram of the DC capacitor voltage waveform of the photovoltaic power station when the grid frequency drops. Figure 5 (b) is a waveform diagram of the active power output of the photovoltaic power station when the grid frequency drops during simulation verification. Detailed Implementation

[0038] The present invention will be further described in detail below with reference to the accompanying drawings and specific embodiments.

[0039] like Figure 1 and Figure 2 As shown, the data-driven DC capacitor grid self-synchronizing new energy power station control method of the present invention is as follows:

[0040] First, voltage transformers and current transformers are used to collect the output current, output voltage, and grid-side current information of new energy power plants such as photovoltaic or direct-drive wind turbines, and then the corresponding Park transformations are performed to obtain their respective data. The system also collects DC capacitor voltage information from renewable energy power plants, providing a basis for subsequent control. These power plants include one renewable energy source, such as photovoltaic or direct-drive wind turbines. Figure 2 As shown, taking photovoltaic power stations as an example, new energy power stations This is the DC capacitor voltage. It is a DC capacitor. The output active power of the photovoltaic power station. To output current for photovoltaic power plants This refers to the output voltage of the photovoltaic power station. For grid-side current, For the LCL filter, For grid-side capacitors, For network testing filter inductor, and For transmission line inductance and resistance, This is the grid voltage. It is a modulated wave voltage signal.

[0041] In practical implementation, the voltage and current information of the renewable energy power station is first obtained. All renewable energy units within the power station are aggregated and considered as a single-unit grid-connected system, referred to as the system. By default, the renewable energy power station uses an LCL-type filter as the filtering component, and the filter inductance of the renewable energy power station is... Filter inductor The equivalent resistance is The filter capacitor is The network test filter inductor is The voltage at the grid connection point of the renewable energy power plant is collected using a voltage transformer as the output voltage. The filter inductance of the new energy power station is collected through a current transformer. The current is used as the output current. The grid-side filter inductance is collected through a current transformer. The current is used as the grid-side current. Obtain the DC capacitor voltage. .

[0042] Then, the output current, output voltage, and grid-side current of the renewable energy power station are respectively subjected to Park transformation to obtain the corresponding values. The axis components are as follows:

[0043]

[0044]

[0045]

[0046] in, To control the frequency; , and The output current of the new energy power station The currents of phases A, B, and C, and These are the output currents of the new energy power stations. Axial components; , and The output voltage of the new energy power station The voltages of phases A, B, and C, and These are the output voltages of the new energy power stations. Axial components; , and These are the grid-side currents of the new energy power plants. The currents of phases A, B, and C, and These are the grid-side currents of the new energy power plants. Axial components.

[0047] Then, a DC capacitor optimal self-synchronization unit for the new energy power station, including a data-driven DeePC controller, is constructed. The specific details of the data-driven DeePC controller are as follows:

[0048] ;

[0049] in, As decision variables, , Represents the set of real numbers. Decision variables Dimensions , , and These are historical data collection duration, adjacent data collection duration, and predicted duration, respectively. , and All are positive integers; and Input and output slack variables are used respectively to enhance system robustness; and These are the inputs and outputs to be optimized for a stand-alone grid-connected system; and These are the input and output constraints of a stand-alone grid-connected system, namely, the preset upper and lower limits. , m and p are the system input and output dimensions, respectively. , ; For vectors quadratic form , =u or yr, P is the cost matrix, P=R or Q, R and Q are positive definite matrices and positive semi-definite matrices respectively, that is, the cost matrices of system input and system output; The second norm of vector S The square of S = or , = ; This serves as the reference output trajectory for a stand-alone grid-connected system. ; , and They are respectively , and The scaling factor, For coefficient The scaling regularization term is usually... square of the 2 norm To improve the closed-loop stability of the system; It is a column vector; and These represent the first input and first output trajectories of a single-unit grid-connected system for predicting historical time periods. and These are the second input and second output trajectories of a single-unit grid-connected system for a predicted historical time period, with a duration of [duration missing]. , , , , ; and Each is the previous time of the current moment Input and output trajectories of a stand-alone grid-connected system collected over a period of time. , .

[0050] The input to a stand-alone grid-connected system includes the actual value of the DC capacitor voltage. Compared with reference value difference of squares and the q-axis component of the output voltage of the new energy power station The output includes the deviation of the control frequency signal. The single-unit grid-connected system is obtained by aggregating the equivalent value of new energy units, and the output of the DeePC controller is used as the input of the system.

[0051] When the DeePC controller generates the control frequency signal, in At time, the collection time sequence length is The most recent input and output trajectories are as follows:

[0052]

[0053] Construct the most recent system input trajectory and system output trajectory vectors ,as follows:

[0054]

[0055] Based on behavioral systems theory, the optimization problem is solved in a rolling manner to obtain the optimal control sequence. .

[0056] The data-driven DeePC controller also requires pre-training, collecting system input and output data for a preset historical time period and constructing a Hankel matrix: injecting white noise perturbation signals into the controller input to ensure that the controller input is a continuous excitation signal, and collecting time-series data with a duration of [missing information]. System input trajectory and system output trajectory ,remember and They are respectively in System inputs and outputs at specific times. , ,as follows:

[0057]

[0058] Construct separately and of Hankel matrix of order And perform the block division operation as follows:

[0059]

[0060]

[0061] Hankel matrix of order It is by and It was constructed.

[0062] The inputs and outputs to be optimized for a stand-alone grid-connected system at time t are as follows:

[0063] ;

[0064]

[0065]

[0066] ;

[0067] ;

[0068] in, and The DeePC controller is a data-driven controller. Input and output at any moment For the DeePC controller in The first output at any given moment and The DeePC controller is a data-driven controller. The first and second inputs at time 1; for Constantly control the deviation of the frequency signal; for The squared difference between the actual value and the reference value of the DC capacitor voltage at any given time. and The DC capacitor voltage is respectively at The actual value and reference value at that moment For the output voltage of the new energy power station Moment Axial components.

[0069] The squared difference between the DC capacitor voltage of the new energy power station and the reference value, as well as the q-axis component of the output voltage, are input into the DC capacitor optimal self-synchronization unit. The DC capacitor optimal self-synchronization unit then drives the control frequency signal output by the DeePC controller based on this data. The modulation wave voltage phase signal is obtained by performing integration. , The data-driven DeePC controller employs a rolling optimization strategy, using rolling solutions to obtain... Input of a stand-alone grid-connected system that is optimized in real time That is, the deviation of the control frequency signal. , To control the time domain, the signal is then input into a stand-alone grid-connected system to obtain the control frequency signal. , , The current time is the rated frequency of the power grid. Update to the next sampling time The re-acquisition timing length is The latest system input and output trajectories are used to update the column vector. The optimization problem is then solved again to obtain the control frequency. In other words, this rolling optimization strategy achieves optimal self-synchronization of the DC capacitors in the renewable energy power plant.

[0070] Then, the output current, output voltage, and grid-side current of the new energy power station are... The axial component is controlled by a dual closed-loop system of voltage and current, which includes both inner-loop current control and inner-loop current control, to obtain the modulated wave voltage amplitude signal. Axis component. Specifically, this refers to the output voltage and its reference value of the renewable energy power station, as well as the grid-side current. The axial component is controlled by the outer voltage loop to obtain current limiting, thereby obtaining the output current. The d-axis component reference value, together with the output current of the new energy power station, is used to obtain the d-axis component U of the modulated wave voltage amplitude signal after passing through the inner current loop control. td and q-axis component U tq .

[0071] 1) Voltage outer loop control:

[0072] The filter capacitor is established based on Kirchhoff's Voltage Law (KVL). The electrical model is as follows:

[0073]

[0074] in, For filtering capacitors, and These are the output current and output voltage of the new energy power station, respectively. This represents the grid-side current.

[0075] For filter capacitors The electrical model is subjected to Park transform and Laplace transform in sequence to obtain the filter capacitor. The frequency domain model is as follows:

[0076]

[0077] in, and For the output current of new energy power stations Axial components; For the Laplace operator; and For grid-side current Axial components.

[0078] Combined with filter capacitor The frequency domain model uses a PI controller in the voltage outer loop control to control the d-axis component of the output voltage of the renewable energy power station. and q-axis components The output of the voltage outer loop control is:

[0079]

[0080] in, and The output is controlled by the outer voltage loop. and These are the output voltages of the outer voltage loop. shaft and Reference value for the axis. and For the output voltage of the converter shaft and Axial components, and These are the proportional and integral coefficients of the voltage outer loop PI controller, respectively. This represents the grid-side current feedforward coefficient.

[0081] 2) Reference for generating current inner loop control:

[0082] Output of voltage outer loop control and Current limiting is performed to obtain the reference for the inner current loop:

[0083]

[0084] in, and These are the current dq-axis references for the inner current loop, respectively. This is the current limiting value; when the output of the outer voltage loop control is less than or equal to... When the current limiting circuit is active, the current limiting circuit does not function; however, when the output of the outer voltage loop control is greater than... The reference current amplitude of the inner current loop is then limited to... Furthermore, priority is given to increasing the d-axis current to meet the active power output requirements of the new energy power station.

[0085] 3) The modulation wave voltage amplitude signal is provided by the inner current loop control:

[0086] Establish the filter inductance based on Kirchhoff's current law. The electrical model is as follows:

[0087]

[0088] in, Filter inductor for new energy power plants, filter inductor The equivalent resistance is , and These are the output current and output voltage of the new energy power station, respectively. It is a modulated wave voltage signal.

[0089] For filter inductors Performing Park and Laplace transforms on the electrical model yields the filter inductor. The frequency domain model is as follows:

[0090]

[0091] Combined with filter inductor The frequency domain model is used, and a PI controller is employed in the inner current loop control to control the d-axis component of the output current of the renewable energy power station. and q-axis components The modulation wave voltage amplitude signal is given by the inner current loop control as follows:

[0092]

[0093] in, and These are the proportional and integral coefficients of the current inner-loop PI controller.

[0094] Then, it is necessary to configure the reference values ​​for the output voltage of the new energy power station, the reference values ​​for the DC capacitor voltage, the grid rated frequency, the DeePC controller parameters, the PI controller parameters for the current inner loop control, and the PI controller parameters for the voltage outer loop control, to complete the parameter settings for the data-driven network-based optimal self-synchronization control of the DC capacitor. When configuring the parameters, configure the reference value for the output voltage of the new energy power station. and DC capacitor voltage reference value Rated frequency value of power grid DeePC controller parameters: timing length of system input and output trajectories. The time series length of the system's most recent input and output trajectories. Predicted length Control Time Domain The reference trajectory output by the system System input cost matrix and system output cost matrix Slack variable weights and Scaling factor of regularization term Parameters of the PI controller for current inner loop control: and Parameters of the PI controller for voltage outer loop control: and The parameters of the PI controller must meet the requirements of system stability and speed. By configuring the above parameters, the parameter settings for data-driven, network-based optimal self-synchronization control of DC capacitors are completed.

[0095] Finally, based on the modulated wave voltage phase signal and the modulated wave voltage amplitude signal... The modulated wave voltage signal obtained from the axis component is transformed and modulated before being applied to the power switching devices of the converter in the renewable energy power station to achieve power station control. Specifically, based on the modulated wave voltage phase signal... The d-axis component U of the modulated wave voltage amplitude signal td and q-axis component U tq After performing the inverse Park transform, the modulated wave voltage signal U is obtained. tabc Then, a pulse width modulation signal is generated by a space vector pulse width modulation (SVPWM) generator and applied to the power switching devices of the converter in the new energy power station.

[0096] 1) Generation of modulated wave voltage signal:

[0097] For modulated wave voltage phase signal and modulated wave voltage amplitude signal , Performing the inverse Park transform yields the modulated voltage signal, as follows:

[0098]

[0099] in, , and For modulated wave voltage signal U tabc The A-phase, B-phase, and C-phase components, and These are respectively the modulated wave voltage signals shaft and Axial components.

[0100] 2) Generation of pulse width modulation signals:

[0101] The obtained modulated voltage signal is passed through an SVPWM generator to obtain a pulse width modulation signal. The pulse width modulation signal is then applied to the power switching devices of the converter to realize data-driven network-type DC capacitor optimal self-synchronization new energy power station control.

[0102] Specific embodiments of the present invention are as follows:

[0103] like Figure 3 As shown, in a specific embodiment of the present invention, a photovoltaic power station is selected as a representative of new energy power stations. This embodiment is a grid-connected simulation system for a photovoltaic power station, and adopts grid following (GFL) control, DC capacitor voltage optimal self-synchronizing control (Data-Driven Self-Synchronizing 1, DDSS1), and the data-driven grid-type DC capacitor optimal self-synchronizing control (Data-Driven Self-Synchronizing 2, DDSS2) of the present invention as the control strategies for the photovoltaic power station. Specifically, the input of the DeePC controller in DDSS1 is only the square difference between the actual value and the reference value of the DC capacitor voltage, while the input of the DeePC controller in DDSS1 includes: the square difference between the actual value and the reference value of the DC capacitor voltage, and the q-axis component of the output voltage of the photovoltaic power station. The relevant simulation parameter settings are shown in Table 1.

[0104] Table 1

[0105]

[0106] Reference value of output voltage of new energy power station and DC capacitor voltage reference value Rated frequency value of power grid DeePC controller parameters: timing length of system input and output trajectories. The time series length of the system's most recent input and output trajectories. Predicted length Control Time Domain The reference trajectory output by the system System input cost matrix and system output cost matrix Slack variable weights and Scaling factor of regularization term , This is the Kronecker product. Parameters of the PI controller for current inner loop control: and Parameters of the PI controller for voltage outer loop control: and When DDSS1 is used as the control strategy, When DDSS2 is used as the control strategy, . Indicates that the diagonal element is and diagonal matrix, Indicates the order is The identity matrix.

[0107] Short-circuit ratio of photovoltaic power plant grid-connected system The active power output of the photovoltaic power station jumps from 0.2 pu to 0.5 pu at t = 0.2 s. The simulation is run and the DC capacitor voltage and output power of the photovoltaic power station grid-connected system are recorded as follows: Figure 4 (a) and Figure 4As shown in (b), it can be seen that the data-driven grid-based DC capacitor optimal self-synchronization control is more adaptable to weak power grids than the traditional grid-following control, and it does not exhibit stability problems. Under weak power grid conditions, the traditional grid-following control GFL system exhibits obvious power oscillations, with continuous fluctuations in the output power of the photovoltaic power station. This reflects an unfavorable interaction between the system and the grid impedance under low short-circuit ratio conditions, resulting in poor power angle stability and insufficient adaptability to weak power grids. In contrast, DDSS1 and DDSS2 do not exhibit power oscillations; their power curves are stable and their response is rapid, demonstrating excellent transient and steady-state performance. This result indicates that the DDSS1 and DDSS2 control methods can effectively avoid the inherent stability problems of traditional phase-locked loop-based GFL control, exhibiting stronger robustness and better dynamic response characteristics in weak power grid environments, providing an effective stability solution for the integration of new energy power stations into weak power grids.

[0108] To further verify that the method of this invention can provide support to the power grid and has networking characteristics, the power grid frequency was set to drop by 0.5Hz at t=0.2s. The simulation model was run and the DC capacitor voltage and output power of the photovoltaic power station grid-connected system were recorded as follows: Figure 5 (a) and Figure 5 As shown in (b), it can be seen that when the grid frequency drops, the traditional grid-connected control system has weak frequency support capability, with slow power transmission response and limited regulation capability, reflecting its inability to provide effective active power support under frequency disturbances. In contrast, DDSS1 and DDSS2 exhibit significant frequency support characteristics: during frequency drops, both can respond quickly, supporting the grid frequency by adjusting active power output, demonstrating good voltage source behavior and autonomous frequency regulation capability. This result indicates that the DDSS1 and DDSS2 control strategies can enable new energy power plants to exhibit grid operation characteristics similar to synchronous generators, not only enhancing the dynamic support for grid frequency but also significantly improving system stability and operational robustness, making them suitable for the stable operation requirements of new power systems.

[0109] This invention also designs a data-driven DC capacitor grid self-synchronization new energy power station control system. The system includes a unit processing module, a dual closed-loop control module, and a power station control module. The unit processing module is used to construct an optimal self-synchronization unit for the DC capacitors of the new energy power station, including a data-driven DeePC controller. It inputs the squared difference between the DC capacitor voltage and the reference value of the new energy power station, as well as the q-axis component of the output voltage, into the optimal self-synchronization unit, processes it, and outputs a modulated wave voltage phase signal. The dual closed-loop control module is used to control the output current, output voltage, and grid-side current of the new energy power station. The axial component is controlled by a dual closed-loop system of voltage and current, which includes both inner-loop current control and inner-loop current control, to obtain the modulated wave voltage amplitude signal. Axis component. The station control module is used to determine the phase signal and amplitude signal of the modulated wave voltage. The shaft component obtains the modulated wave voltage signal, which is then transformed and modulated before being applied to the power switching devices of the converter in the new energy power station to realize the control of the new energy power station.

[0110] The above are merely preferred embodiments of the present invention and do not constitute any limitation on the present invention. Any equivalent substitutions or modifications made by those skilled in the art to the technical solutions and content disclosed in the present invention without departing from the scope of the present invention shall be deemed to have remained within the protection scope of the present invention.

Claims

1. A data-driven control method for self-synchronizing DC capacitor-connected renewable energy power stations, characterized in that, include: Step 1) Construct a DC capacitor optimal self-synchronization unit for the new energy power station, including a data-driven DeePC controller. Input the square difference between the DC capacitor voltage of the new energy power station and the reference value, as well as the q-axis component of the output voltage, into the DC capacitor optimal self-synchronization unit. After processing, output the modulated wave voltage phase signal. Step 2) Measure the output current, output voltage, and grid-side current of the renewable energy power station. The axial component is controlled by a dual closed-loop system of voltage and current, which includes both inner-loop current control and inner-loop current control, to obtain the modulated wave voltage amplitude signal. Axial components; Step 3) Based on the modulation wave voltage phase signal and the modulation wave voltage amplitude signal... The shaft component obtains the modulated wave voltage signal, which is then transformed and modulated before being applied to the power switching devices of the converter in the new energy power station to realize the control of the new energy power station. In step 1), the data-driven DeePC controller is specifically as follows: ； in, For decision variables; and These are the input and output slack variables, respectively. and These are the inputs and outputs to be optimized for a stand-alone grid-connected system; and These are the input and output constraints for a stand-alone grid-connected system, respectively. For vectors quadratic form , =u or yr, P is the cost matrix, P=R or Q, R and Q are positive definite and positive semi definite matrices respectively; The second norm of vector S The square of S = or , = ; This serves as the reference output trajectory for a stand-alone grid-connected system. , and They are respectively , and The scaling factor, For regularization terms; It is a column vector; and These represent the first input and first output trajectories of a single-unit grid-connected system for predicting historical time periods. and These are the second input and second output trajectories of a single-unit grid-connected system for a predicted historical time period, with a duration of [duration missing]. ; and Each is the previous time of the current moment Input and output trajectories of a stand-alone grid-connected system collected over a period of time; The input to a stand-alone grid-connected system includes the actual value of the DC capacitor voltage. Compared with reference value difference of squares and the q-axis component of the output voltage of the new energy power station The output includes the deviation of the control frequency signal. The single-unit grid-connected system is obtained by aggregating the equivalent value of new energy units.

2. The data-driven DC capacitor grid self-synchronizing new energy power station control method according to claim 1, characterized in that: The specific inputs and outputs to be optimized for the single-machine grid-connected system at time t are as follows: ； ； ； in, and The DeePC controller is a data-driven controller. Input and output at any moment and The DeePC controller is a data-driven controller. The first and second inputs at time 1; for Constantly control the deviation of the frequency signal; for The squared difference between the actual value and the reference value of the DC capacitor voltage at any given time. and The DC capacitor voltage is respectively at The actual value and reference value at that moment For the output voltage of the new energy power station Moment Axial components.

3. The data-driven DC capacitor grid self-synchronizing new energy power station control method according to claim 1, characterized in that: The data-driven DeePC controller employs a rolling optimization strategy, using rolling solutions to obtain... Input of a stand-alone grid-connected system that is constantly optimized That is, the deviation of the control frequency signal. , To control the time domain, the signal is then input into a stand-alone grid-connected system to obtain the control frequency signal. , , This is the rated frequency of the power grid.

4. The data-driven DC capacitor grid self-synchronizing new energy power station control method according to claim 1, characterized in that: In step 1), the DC capacitor optimal self-synchronization unit drives the control frequency signal output by the DeePC controller based on the data. The modulation wave voltage phase signal is obtained by performing integration. .

5. The data-driven DC capacitor grid self-synchronizing new energy power station control method according to claim 1, characterized in that: In step 2), the output voltage and its reference value of the new energy power station, as well as the grid-side current, are... The axial component is controlled by the outer voltage loop to obtain current limiting, thereby obtaining the output current. The d-axis component reference value, together with the output current of the new energy power station, is used to obtain the d-axis component U of the modulated wave voltage amplitude signal after passing through the inner current loop control. td and q-axis component U tq .

6. The data-driven DC capacitor grid self-synchronizing new energy power station control method according to claim 1, characterized in that: In step 3), based on the modulated wave voltage phase signal The d-axis component U of the modulated wave voltage amplitude signal td and q-axis component U tq After performing the inverse Park transform, the modulated wave voltage signal U is obtained. tabc Then, a pulse width modulation signal is generated by a space vector pulse width modulation (SVPWM) generator and applied to the power switching devices of the converter in the new energy power station.

7. A data-driven DC capacitor grid self-synchronizing new energy power station control system, characterized in that, include: The unit processing module is used to construct the DC capacitor optimal self-synchronization unit of the new energy power station, which includes a data-driven DeePC controller. It inputs the square difference between the DC capacitor voltage and the reference value of the new energy power station and the q-axis component of the output voltage into the DC capacitor optimal self-synchronization unit, and outputs the modulated wave voltage phase signal after processing. The dual closed-loop control module is used to control the output current, output voltage, and grid-side current of the renewable energy power station. The axial component is controlled by a dual closed-loop system of voltage and current, which includes both inner-loop current control and inner-loop current control, to obtain the modulated wave voltage amplitude signal. Axial components; The station control module is used to determine the modulation wave voltage phase signal and the modulation wave voltage amplitude signal. The shaft component obtains the modulated wave voltage signal, which is then transformed and modulated before being applied to the power switching devices of the converter in the new energy power station to realize the control of the new energy power station.

8. An electronic device, characterized in that, include: A memory and a processor are coupled to each other, wherein the memory stores program data, and the processor invokes the program data to perform the method as described in any one of claims 1-6.

9. A computer-readable storage medium storing program data thereon, characterized in that, When the program data is executed by the processor, it implements the method as described in any one of claims 1-6.

Images