A deviation evaluation adaptive follow-net and network construction combined control method

Abstract

The application discloses a kind of based on deviation evaluation adaptive grid-connected and grid-constructing fusion control method, belong to grid-connected inverter control technical field.Firstly, the direct current side voltage of grid-connected inverter, filter capacitor voltage, inverter side current and grid phase angle are collected in real time, respectively generate reference instruction under grid-connected type and grid-constructing type mode, then the deviation function value and candidate switch state of two modes are calculated by corresponding model predictive control submodule, after generating basic distribution signal by time slice scheduling distribution module, the time slice proportion of grid-connected and grid-constructing mode is adaptively adjusted according to the numerical relationship of deviation evaluation function, finally, two-way reference instruction is fused and error is corrected, and the optimal control instruction is output to power switch device, while feedback operating parameter realizes closed-loop adaptive control.

Description

Technical Field

[0001] This invention relates to the field of grid-connected inverter control technology, and in particular to a grid-connected and grid-building integrated control method based on deviation assessment and adaptive control. Background Technology

[0002] With the advancement of the global energy transition, renewable energy sources, represented by solar and wind power, are being connected to the grid on a large scale. As the core equipment for power conversion, the control strategy of grid-connected inverters is crucial to the stable operation of the power grid. Traditional grid-connected inverters mainly employ a grid-following (GFL) control strategy, using a phase-locked loop (PLL) to track the grid voltage vector and achieve high-efficiency power transmission. However, in power systems with a high proportion of renewable energy integration, the grid short-circuit ratio (SCR) fluctuates significantly, and the system exhibits significant weak grid characteristics. In a weak grid environment, traditional GFL control faces the risks of resonance and instability, making it difficult to guarantee current quality.

[0003] To improve stability in weak power grid environments, grid-forming (GFM) control has gradually gained attention. GFM control possesses the ability to actively support voltage and frequency, exhibiting strong stability in weak power grid environments. However, GFM control still has shortcomings in practical applications: on the one hand, in strong power grid environments, the virtual inertia it introduces may lead to a second-order oscillatory element, generating low-frequency oscillations that are difficult to converge for extended periods; on the other hand, during large disturbances or voltage drops, GFM mode is highly susceptible to generating instantaneous inrush currents, endangering the safety of power devices.

[0004] In summary, neither a single grid-following nor grid-building control mode can simultaneously meet the requirements of stability, dynamic response, and power quality in complex environments with large fluctuations in grid impedance (SCR).

[0005] Currently, several academic papers and patents have analyzed and reported on the inability of single-grid-connected and single-grid-structured control systems to maintain stable operation of grid-connected inverters under conditions of significant fluctuations in grid impedance (SCR), for example: 1. Patent application CN2023109669157, entitled "A Fusion Control Method for Grid-Connected Inverters Based on Time-Sharing of Voltage and Current Sources," discloses a fusion control method for grid-connected inverters based on time-sharing of voltage and current sources. This patent addresses the problem of stable operation of grid-connected inverters under significant fluctuations in the grid short-circuit ratio. It proposes dividing each control process into N+M control cycles and using a counting method to achieve time-sharing switching between voltage source mode (N cycles) and current source mode (M cycles). This method integrates the characteristics of both modes to improve the stability of the inverter under wide operating conditions, attempting to solve the adaptability defects of a single mode under strong and weak grid conditions. However, this technology still has limitations in engineering applications. The patent uses a fixed-count time-sharing switching strategy, with N and M being preset fixed values. It cannot dynamically adjust the proportion of the two modes according to the real-time fluctuations in grid impedance. When the grid short-circuit ratio changes abruptly, the fixed cycle allocation is difficult to quickly adapt to changes in operating conditions, easily leading to voltage and current fluctuations caused by mode switching lag.

[0006] 2. The paper titled "The Dual-Mode Combined Control Strategy for Centralized Photovoltaic Grid-Connected Inverters Based on Double-Split Transformers" by M.Li, X. Zhang, Z. Guo, J. Wang, and F. Li proposes a dual-mode control approach for multi-inverter systems. Based on different grid impedances, the system operates in three different control modes: grid-connected control, grid-connected control, and integrated control. This improves inverter stability under significant SCR fluctuations. However, grid impedance identification technology is currently immature, resulting in poor engineering applicability. Summary of the Invention

[0007] To overcome the shortcomings of the prior art, the present invention provides a grid-connected and grid-building integrated control method based on deviation assessment, which solves the problems of poor stability of existing grid-connected control strategies when grid impedance fluctuates drastically and the excessive reliance of integrated control on impedance identification.

[0008] To achieve the above objectives, the present invention adopts the following technical solution, including: A method for adaptive network following and network construction fusion control based on deviation assessment includes the following steps: Step 1: Real-time acquisition of DC-side voltage and filter capacitor voltage of the grid-connected inverter. v c1_abc Inverter-side current i L1_abc Phase-locked loop generates power grid phase angle grid _ wt ; Step 2, preset active power P and reactive power Q Based on root mean square feedback regulation, the current reference command in the grid mode is generated through the current reference calculation module. i ref_d , i ref_q Preset active power P reactive power Q Combined with the sampled filter capacitor voltage v c1_abc Inverter-side current i L1_abc The voltage reference calculation module generates voltage reference instructions in the network configuration mode. v ref_d , v ref_q ; Step 3, Current Reference Command i ref_d , i ref_q Generate after coordinate transformation , The inverter-side current collected in step 1 i L1_abc Generate after coordinate transformation , ;Will , and , The input is fed into the network-following model predictive control submodule. Based on the preset network-following control logic and combined with the real-time collected system operating data, the network-following model predictive control submodule calculates the deviation function value under the network-following mode. J GFL and candidate switch status; Step 4, Voltage Reference Command v ref_d , v ref_q Generate after coordinate transformation , The voltage of the filter capacitor collected in step 1 v c1_abc Generate after coordinate transformation , ;Will , and , The input network-based model predictive control submodule calculates the deviation function value under the network-based mode based on preset network control logic and real-time collected system operating data. JGFM and candidate switch status; Step 5, the deviation function values ​​generated in steps 3 and 4 J GFL and J GFM The input is sent to the time slice scheduling and allocation module, which generates a basic time slice allocation signal based on the preset scheduling logic. This basic time slice allocation signal represents the initial time slice ratio benchmark of the network connection and network construction modes. Step 6: Based on the calculation logic of the deviation evaluation function, the control mode with higher cost has a lower allocated time slice percentage; by comparison... J GFL and J GFM The numerical relationship is used to adjust the output time slice mode signal in real time, so as to achieve adaptive correction of the time slice ratio of the network following mode and the network construction mode. Step 7: Based on the weighting determined by the time slice mode signal, perform fusion calculation on the reference instructions of the two sub-modules, and at the same time, combine the cost value to correct the error in the fusion process; finally, output the optimal control instruction after fusion through the switch quantity, and feed back the cost evaluation result to the system to complete the calculation of a single control cycle. Step 8: Apply the optimal control command output in step 7 to the power switching devices of the grid-connected inverter to achieve closed-loop regulation of output voltage and current; at the same time, provide real-time feedback on the inverter's operating parameters, repeat steps 1 to 7 to achieve closed-loop adaptive control and continuously adapt to the fluctuations in the grid short-circuit ratio.

[0009] Preferably, in step 2, the current reference command in the grid mode is followed. i ref_d , i ref_q Voltage reference command in grid-type mode v ref_d , v ref_q The calculation formula is as follows:

[0010]

[0011] in, The given active power reference value; The given reactive power reference value; This represents the amplitude of the filter capacitor voltage. and These are the feedback values ​​for active power and reactive power, respectively. This is the active frequency droop factor; This is the reactive voltage droop factor; The rated angular frequency; For the Laplace operator; This is the rated voltage amplitude; It outputs reactive power to the inverter.

[0012] Preferably, in step 3, the deviation function value in the net mode J GFL The formula is as follows:

[0013] In the formula, n For current tracking weighting coefficients; i ( k - n +1) is the current reference value; i ( kn This is the predicted current value from the previous moment; n To predict the time-domain order; k Indicates the current control moment.

[0014] Preferably, the current reference value is generated based on the MPPT algorithm, which is the maximum power point tracking algorithm, and calculates the current reference value based on the given active power and reactive power.

[0015] Preferably, in step 4, the deviation function value under the meshing pattern J GFM The formula is as follows:

[0016] In the formula, n For current tracking weighting coefficients; v ( kn +1) is the reference value for the rated voltage; v ( kn This is the predicted value of the voltage at the previous moment; n To predict the time-domain order; k Indicates the current control moment.

[0017] Preferably, in step 6, the guiding principle for time partitioning is as follows:

[0018]

[0019] In the formula, This determines the time allocation for network connection and network construction; z It is the threshold of the deviation cost coefficient; GFM represents the network type; GFL represents the following network type.

[0020] The present invention also provides a readable storage medium having a computer program stored thereon, which, when executed, implements the aforementioned adaptive network-following and network-building fusion control method based on deviation evaluation.

[0021] The present invention also provides an electronic device, which includes a processor, a memory, and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, it implements the aforementioned adaptive network following and network construction fusion control method based on deviation evaluation.

[0022] The present invention also provides a computer program product comprising a computer program / instruction that, when executed by a processor, implements the aforementioned adaptive network-following and network-building fusion control method based on deviation evaluation.

[0023] The advantages of this invention are: (1) This invention does not require online identification of grid impedance and can dynamically and smoothly adjust the time slice ratio of the two control modes. Without increasing the system control order, it can achieve stable operation of grid-connected inverters in scenarios with large fluctuations in grid impedance. It has good dynamic response characteristics and power quality, and the parameter design is simple and has strong engineering applicability.

[0024] (2) This invention is a model predictive control method that integrates grid-following (GFL) and grid-forming (GFM) modes by adaptively adjusting the time slice allocation ratio under the background of wide-range grid impedance fluctuations.

[0025] (3) This invention proposes an adaptive time-slice model predictive control (MPC) strategy based on deviation evaluation function. By backtracking past state moments, the prediction effect of the predictive system is evaluated. According to the deviation evaluation function, the time slice ratio of grid-connected control and grid-connected control is adaptively allocated. When the grid impedance fluctuates greatly, the grid-connected inverter system can maintain stable operation and fast response. Attached Figure Description

[0026] Figure 1 This is a topology diagram of a new energy converter grid-connected system according to an embodiment of the present invention.

[0027] Figure 2 This is a block diagram of the fusion control structure for the network and the network structure of the present invention.

[0028] Figure 3 The output current waveforms and THD analysis diagrams for fusion control, single grid control, and single grid construction control in embodiments of the present invention at SCR=50 are shown.

[0029] Figure 4The output current waveforms and THD analysis diagrams for fusion control, single grid control, and single grid construction control in embodiments of the present invention at SCR=3 are shown.

[0030] Figure 5 The output current waveforms and THD analysis diagrams for fusion control, single grid control, and single grid construction control in embodiments of the present invention at SCR=1.5 are shown. Detailed Implementation

[0031] 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 scope of protection of the present invention.

[0032] Figure 1 This is a topology diagram of a new energy converter grid-connected system according to an embodiment of the present invention. Figure 1 The DC side provides new energy DC power from sources such as photovoltaics, wind power, or energy storage, with a DC voltage of [missing information]. v dc The rated phase voltage of the three-phase AC output is 220V, which is 800V. V / 50 Hz .

[0033] Figure 2 This is a diagram of the network-following and network-building fusion control structure based on the model prediction adaptive time slice using the deviation evaluation function, according to an embodiment of the present invention. Figure 2 As can be seen, the present invention provides an adaptive control method for network following and network construction based on deviation evaluation, comprising the following steps: Step 1: Real-time acquisition of DC-side voltage and filter capacitor voltage of the grid-connected inverter. v c1_abc Inverter-side current i L1_abc Phase-locked loop (PLL) generates the power grid phase angle. grid _ wt .

[0034] Among them, the phase angle of the power grid generated by the phase-locked loop grid _ wt , used as a phase reference for generating current reference commands.

[0035] Step 2: Preset active power P and reactive power Q, adjust based on root mean square (RMS) feedback, and generate current reference commands in grid-following (GFL) mode through the current reference calculation module. i ref_d , i ref_qSimilarly, the preset active power P and reactive power Q are combined with the sampled filter capacitor voltage. v c1_abc Inverter-side current i L1_abc The voltage reference calculation module generates voltage reference commands in the grid-mode (GFM) configuration. v ref_d , v ref_q The relevant calculation formulas are as follows:

[0036]

[0037] in, For a given active power reference value, this embodiment is 500kW; Given a reactive power reference value, this embodiment sets it to 0; This is the amplitude of the filter capacitor voltage, specifically a value obtained through real-time sampling and calculation. and These are the feedback values ​​for active power and reactive power, respectively. This is the active frequency droop factor; This is the reactive voltage droop factor; This is the rated angular frequency; 50 Hz corresponds to an angular frequency of 100 π rad / s. For the Laplace operator; This is the rated voltage amplitude; It outputs reactive power to the inverter.

[0038] Step 3, Current Reference Command i ref_d , i ref_q After coordinate transformation ( After generating , The inverter-side current collected in step 1 i L1_abc After coordinate transformation ( After generating , ;Will , and , The input follows the predictive control submodule of the network-following model. This module calculates the deviation function value under the network-following mode based on the preset network-following control logic and the real-time collected system operating parameters. J GFL and candidate switch states, deviation function values ​​in grid mode. J GFLThe formula is as follows:

[0039] This formula is used to calculate the error between the predicted current value and the actual reference value over the past three time points, and is used to generate a deviation function value, which serves as a criterion for evaluating the prediction performance of the prediction system.

[0040] In the formula, n The current tracking weighting coefficient reflects the magnitude of the influence of different past moments on the current moment's state; i ( k - n +1) is the current reference value generated based on the MPPT algorithm; i ( kn This is the predicted value of the current at the previous moment; n To predict the time-domain order; k This indicates the current control moment; the MPPT algorithm is the maximum power point tracking algorithm, used to generate current reference values, which are calculated based on the given active and reactive power.

[0041] Step 4, Voltage Reference Command v ref_d , v ref_q After coordinate transformation ( After generating , The voltage of the filter capacitor collected in step 1 v c1_abc After coordinate transformation ( After generating , ;Will , and , The input network model predictive control submodule calculates the deviation function value under the network mode based on preset network control logic and real-time collected system operating data. J GFM and candidate switch states, deviation function values ​​in network configuration mode. J GFM The formula is as follows:

[0042] In the formula, n The current tracking weighting coefficient reflects the magnitude of the influence of different past moments on the current state. v ( kn +1) is the reference value for the rated voltage. v( kn This is the predicted value of the voltage at the previous moment.

[0043] Step 5, the deviation function values ​​generated in steps 3 and 4 J GFL and J GFM The input is sent to the time slice scheduling and allocation module, which generates a basic time slice allocation signal based on preset scheduling logic. time _ slice This signal characterizes the initial time slice ratio of the network and network construction modes, providing an initial reference framework for subsequent dynamic adjustments.

[0044] Step 6: Based on the calculation logic of the deviation evaluation function, the control mode with higher cost has a lower allocated time slice percentage; by comparison... J GFL and J GFM The numerical relationship is used to adjust the output time-slice mode signal in real time, achieving adaptive correction of the time-slice ratio of network following and network construction modes, and avoiding control defects caused by long-term dominance of a single mode. The guiding principle for time-slicing is as follows:

[0045]

[0046] In the formula, The time allocation for network connection and network construction is determined. z It is the threshold of the deviation cost coefficient.

[0047] Step 7: Based on the weighting determined by the time slice mode signal, perform fusion calculation on the reference instructions of the two sub-modules, and at the same time, combine the cost value to correct the error in the fusion process; finally, output the fused optimal control instruction through a switch, and feed back the cost evaluation result to the system to complete the calculation of a single control cycle.

[0048] Step 8: Apply the optimal control command output in Step 7 to the power switching devices of the grid-connected inverter to achieve closed-loop regulation of the output voltage / current; at the same time, provide real-time feedback on the inverter's operating parameters and repeat Steps 1 to 7 to achieve closed-loop adaptive control and continuously adapt to the fluctuations in the grid short-circuit ratio.

[0049] This embodiment presents an invention applicable to a model prediction adaptive time-slice and grid / network fusion control method based on a deviation evaluation function under conditions of large fluctuations in grid impedance.

[0050] To demonstrate the technical effects of this invention, Figure 1The simulation was performed on the 500kW three-phase two-level grid-connected converter shown. This converter uses... Figure 2 The model prediction adaptive time slice and network / network fusion control method based on the deviation evaluation function is shown.

[0051] Figure 3 The diagram shows the output current waveforms and THD analysis of the integrated control, single grid-following control, and single grid-building control in this embodiment of the invention at SCR=50. It can be seen that under SCR=50, i.e., under a strong grid, the output current waveform of the single grid-following control is stable, with a THD of 2.72%, meeting the system grid connection requirements; the output current waveform of the single grid-building control is unstable, with a THD of 10.53%, failing to meet the system grid connection requirements; the output current waveform of the integrated control strategy proposed in this invention is stable, with a THD of 2.07%, meeting the system grid connection requirements.

[0052] Figure 4 The diagram shows the output current waveforms and THD analysis of the integrated control, single grid-following control, and single grid-building control in this invention when SCR=3. It can be seen that under SCR=3, i.e., a weak grid, the output current waveform of the single grid-following control is unstable, with a THD of 8.01%, which does not meet the system grid connection requirements; the output current waveform of the single grid-building control is stable, with a THD of 1.92%, also not meeting the system grid connection requirements; the output current waveform of the integrated control strategy proposed in this invention is stable, with a THD of 1.76%, meeting the system grid connection requirements.

[0053] Figure 5 The diagram shows the output current waveforms and THD analysis of the integrated control, single grid-following control, and single grid-building control in this embodiment of the invention at SCR=1.5. It can be seen that under SCR=1.5, i.e., under extremely weak grid conditions, the output current waveform of the single grid-following control is unstable, with a THD of 12.58%, which does not meet the system grid connection requirements; the output current waveform of the single grid-building control is stable, with a THD of 1.03%, also not meeting the system grid connection requirements; the output current waveform of the integrated control strategy proposed in this invention is stable, with a THD of 1.49%, meeting the system grid connection requirements.

[0054] The above are merely preferred embodiments of the present invention and are not intended to limit the scope of the present invention. Any modifications, equivalent substitutions, and improvements made within the spirit and principles of the present invention should be included within the scope of protection of the present invention.

Claims

1. A method for integrated control of network following and network construction based on deviation evaluation and adaptive control, characterized in that, Includes the following steps: Step 1: Real-time acquisition of DC-side voltage and filter capacitor voltage of the grid-connected inverter. v c1_abc Inverter-side current i L1_abc Phase-locked loop generates power grid phase angle grid _ wt ; Step 2, preset active power P and reactive power Q Based on root mean square feedback regulation, the current reference command in the grid mode is generated through the current reference calculation module. i ref_d , i ref_q Preset active power P reactive power Q Combined with the sampled filter capacitor voltage v c1_abc Inverter-side current i L1_abc The voltage reference calculation module generates voltage reference instructions in the network configuration mode. v ref_d , v ref_q ; Step 3, Current Reference Command i ref_d , i ref_q Generate after coordinate transformation , The inverter-side current collected in step 1 i L1_abc Generate after coordinate transformation , ;Will , and , The input is fed into the network-following model predictive control submodule. Based on the preset network-following control logic and combined with the real-time collected system operating data, the network-following model predictive control submodule calculates the deviation function value under the network-following mode. J GFL and candidate switch status; Step 4, Voltage Reference Command v ref_d , v ref_q Generate after coordinate transformation , The voltage of the filter capacitor collected in step 1 v c1_abc Generate after coordinate transformation , ;Will , and , The input network-based model predictive control submodule calculates the deviation function value under the network-based mode based on preset network control logic and real-time collected system operating data. J GFM and candidate switch status; Step 5, the deviation function values ​​generated in steps 3 and 4 J GFL and J GFM The input is sent to the time slice scheduling and allocation module, which generates a basic time slice allocation signal based on the preset scheduling logic. This basic time slice allocation signal represents the initial time slice ratio benchmark of the network connection and network construction modes. Step 6: Based on the calculation logic of the deviation evaluation function, the control mode with higher cost has a lower allocated time slice percentage; by comparison... J GFL and J GFM The numerical relationship is used to adjust the output time slice mode signal in real time, so as to achieve adaptive correction of the time slice ratio of the network following mode and the network construction mode. Step 7: Based on the weighting determined by the time slice mode signal, perform fusion calculation on the reference instructions of the two sub-modules, and at the same time, combine the cost value to correct the error in the fusion process; finally, output the optimal control instruction after fusion through the switch quantity, and feed back the cost evaluation result to the system to complete the calculation of a single control cycle. Step 8: Apply the optimal control command output in step 7 to the power switching devices of the grid-connected inverter to achieve closed-loop regulation of output voltage and current; at the same time, provide real-time feedback on the inverter's operating parameters, repeat steps 1 to 7 to achieve closed-loop adaptive control and continuously adapt to the fluctuations in the grid short-circuit ratio.

2. The adaptive tracking and network construction fusion control method based on deviation evaluation according to claim 1, characterized in that, In step 2, follow the current reference command in mesh mode. i ref_d , i ref_q Voltage reference command in grid-type mode v ref_d , v ref_q The calculation formula is as follows: in, The given active power reference value; The given reactive power reference value; This represents the amplitude of the filter capacitor voltage. and These are the feedback values ​​for active power and reactive power, respectively. This is the active frequency droop factor; This is the reactive voltage droop factor; The rated angular frequency; For the Laplace operator; This is the rated voltage amplitude; It outputs reactive power to the inverter.

3. The adaptive tracking and network construction fusion control method based on deviation evaluation according to claim 1, characterized in that, In step 3, the deviation function value in the following mode J GFL The formula is as follows: In the formula, n For current tracking weighting coefficients; i ( k - n +1) is the current reference value; i ( kn This is the predicted current value from the previous moment; n To predict the time-domain order; k Indicates the current control moment.

4. The adaptive network-following and network-building fusion control method based on deviation evaluation according to claim 3, characterized in that, The current reference value is generated based on the MPPT algorithm, which is the maximum power point tracking algorithm. The MPPT algorithm calculates the current reference value based on the given active power and reactive power.

5. The adaptive tracking and network construction fusion control method based on deviation evaluation according to claim 1, characterized in that, In step 4, the deviation function value under the meshing pattern J GFM The formula is as follows: In the formula, n For current tracking weighting coefficients; v ( kn +1) is the reference value for the rated voltage; v ( kn This is the predicted value of the voltage at the previous moment; n To predict the time-domain order; k Indicates the current control moment.

6. The adaptive network-following and network-building fusion control method based on deviation evaluation according to claim 1, characterized in that, In step 6, the guiding principle for time slicing is as follows: In the formula, This determines the time allocation for network connection and network construction; z It is the threshold of the deviation cost coefficient; GFM represents the network type; GFL represents the following network type.

7. A readable storage medium, characterized in that, It stores a computer program, which, when executed, implements the adaptive network-following and network-building fusion control method based on deviation evaluation as described in any one of claims 1 to 6.

8. An electronic device, characterized in that, It includes a processor, a memory, and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, it implements the adaptive network-following and network-building fusion control method based on deviation evaluation as described in any one of claims 1 to 6.

9. A computer program product, characterized in that, It includes a computer program / instruction that, when executed by a processor, implements the adaptive network-following and network-building fusion control method based on deviation evaluation as described in any one of claims 1 to 6.

Images