Distributed energy intelligent optimization scheduling system and method based on virtual power plant

Abstract

The invention discloses a distributed energy intelligent optimization scheduling system and method based on a virtual power plant, and relates to the technical field of intelligent power grid control. Real-time acquisition and preprocessing are carried out through power grid related data, the accuracy of the data is ensured, then feature extraction is carried out, a new energy disturbance score Rd is obtained, and the operation state of a power grid is evaluated; influences of new energy grid connection on a power grid are sensed in advance, reactive compensation conflicts among inverters are reduced in combination with a conflict matrix C and an iterative optimization algorithm, the precision of reactive scheduling is improved through adaptive compensation control and short-time prediction and interference correction, and finally, the stability of the power grid is monitored in real time, an energy scheduling strategy is optimized, and the power grid stability is improved. And the safety of new energy grid connection and the stability of power grid operation are improved.

Description

Technical Field

[0001] The present invention relates to the field of smart grid control technology, and in particular to a distributed energy intelligent optimization scheduling system and method based on a virtual power plant. Background Art

[0002] In the modern energy management system, smart grid has become an important technical direction to promote global energy transformation. Smart grid uses information and communication technology and advanced power control strategies to achieve supply and demand matching, improve energy utilization efficiency, and adapt to the complex environment of high penetration of new energy. Under the architecture of smart grid, virtual power plant, as a new type of distributed energy aggregation and scheduling platform, can integrate multiple distributed energy sources, including photovoltaic, wind power, energy storage and electric vehicles, to provide flexibility services for the power grid. Especially in the context of the increasing proportion of distributed energy, the distributed energy intelligent optimization scheduling system based on virtual power plant can coordinate multiple energy units to achieve efficient power balance and grid stability. However, in the scenario of high penetration of new energy, VPP scheduling not only needs to manage the power distribution of each distributed energy, but also needs to deal with voltage fluctuations, frequency drift and reactive power regulation brought about by the access of new energy to ensure the safe and stable operation of the power grid.

[0003] In the Chinese invention patent application publication number CN119518978A, a distributed energy management system for smart grids and its optimization scheduling method are disclosed. By integrating advanced data acquisition, processing, optimization scheduling, control execution and user interaction modules, efficient, intelligent and sustainable management of power resources is achieved. The system adopts multi-dimensional optimization scheduling strategy and adaptive control strategy, combined with digital twin technology, to improve the accuracy of real-time monitoring and prediction of grid operation status, enhance the system's response speed and decision-making ability, and can automatically adjust the operating parameters of distributed power generation resources and energy storage systems to achieve precise load control, improve energy utilization efficiency, reduce operating costs, and promote the sustainable development of the power grid. At the same time, the use of user interaction modules improves the transparency of the system and user participation, providing strong technical support for the construction and operation of smart grids.

[0004] The above scheduling methods can optimize the allocation of power resources and enhance the grid's adaptability to emergencies. However, in addition to this, existing energy scheduling optimization usually monitors the stability of power and relies on traditional static reactive power compensation and fixed parameter control strategies.

[0005] However, due to the high volatility and uncertainty of renewable energy output such as photovoltaic and wind power, the grid frequency is prone to drift. The traditional synchronous generator inertia frequency modulation method is difficult to provide sufficient response speed. When multiple distributed energy sources are connected, each inverter independently controls its own active and reactive power. The lack of a global optimization strategy leads to local voltage imbalance, affecting the stable operation of the grid. It is difficult to effectively respond to the dynamic optimization needs of voltage-frequency, resulting in slow system response and low compensation accuracy, affecting the safety and economy of the grid.

[0006] To this end, the present invention provides a distributed energy intelligent optimization scheduling system and method based on a virtual power plant. Summary of the Invention

[0007] In response to the deficiencies in the prior art, the present invention provides a distributed energy intelligent optimization scheduling system and method based on a virtual power plant, which solves the problems in the above-mentioned background technology.

[0008] To achieve the above objectives, the present invention is implemented through the following technical solutions: a distributed energy intelligent optimization and scheduling system based on a virtual power plant, including a data acquisition and preprocessing module, a feature extraction and evaluation module, a collaborative optimization module, an adaptive reactive power compensation control module, and a virtual power plant scheduling evaluation module; The data acquisition and preprocessing module is used to collect power grid related data, perform preprocessing, and construct a power grid related data set S; The feature extraction and evaluation module is used to extract features based on the grid-related data set S, obtain the humidity attenuation coefficient Sd, the local cross interference factor Gr, and the micro-arc interference factor Dh, and obtain the new energy disturbance score Rd to evaluate the grid operation status; The collaborative optimization module is used to calculate the reactive compensation amount of each inverter and generate a reactive compensation plan for the power grid; The adaptive reactive power compensation control module is used to optimize the reactive power compensation scheme according to the power grid to obtain the energy scheduling execution plan; The virtual power plant dispatch evaluation module is used to perform virtual power plant energy dispatch and grid stability evaluation based on the energy dispatch execution plan.

[0009] Preferably, the data acquisition and preprocessing module includes a data acquisition unit and a preprocessing unit; The data acquisition unit is used to determine the power grid topology of the target area and select several power grid nodes. Intelligent sensor groups are deployed at these nodes to collect power grid-related data in real time. Grid nodes include substations, grid connection points, distribution rooms, wind turbines, and photovoltaic power plants. The intelligent sensor groups include synchronized phasor measurement devices, temperature sensors, and humidity sensors. Power grid-related data includes power grid operation data, new energy data, and environmental data. Grid operation data includes voltage V, frequency f, inverter current I, and electromagnetic interference intensity B; New energy data includes wind turbine output power and photovoltaic output power ; Environmental data include temperature T and humidity RH; The preprocessing unit is used to perform noise filtering, outlier removal and data standardization on the collected power grid related data, and to construct a power grid related data set S based on the preprocessed power grid related data.

[0010] Preferably, the feature extraction and evaluation module includes a feature extraction unit and a preliminary evaluation unit; The feature extraction unit is used to extract features based on the grid-related data set S to obtain the humidity attenuation coefficient Sd, the local cross interference factor Gr, and the micro-arcing interference factor Dh. The humidity attenuation coefficient Sd is obtained as follows: ; Where, Indicates the humidity of the target area at the current time point t, Indicates the temperature of the target area at the current time point t, Indicates the historical maximum humidity of the target area, and They represent the historical minimum and maximum temperatures of the target area, respectively, and e represents the base of the natural logarithm. represents the exponential term; The local cross interference factor Gr is obtained as follows: ; Where, Indicates the electromagnetic interference intensity of the target area at the current time point t, and Respectively represent the historical minimum and maximum electromagnetic interference intensity of the target area, represents the output power of wind turbines in the target area at the current time point t, represents the photovoltaic output power of the target area at the current time point t, and They represent the historical maximum wind turbine output power and maximum photovoltaic output power in the target area respectively; The micro-arc interference factor Dh is obtained as follows: ; Where, represents the inverter current value of the target area at the current time point t, Indicates the inverter's power consumption per unit time The current variation amplitude, and They represent the historical maximum current value and minimum current value of the inverter in the target area respectively.

[0011] Preferably, the preliminary evaluation unit is used to perform summary calculation based on the humidity attenuation coefficient Sd, the local cross interference factor Gr, and the micro-arc interference factor Dh to obtain the new energy disturbance score Rd, wherein the new energy disturbance score Rd is obtained as follows: ; The disturbance threshold Rdyz is preset, and the new energy disturbance score Rd and the disturbance threshold Rdyz are compared and analyzed to evaluate the grid operation status. The specific evaluation contents are as follows: If the new energy disturbance score Rd is less than the disturbance threshold Rdyz, that is, Rd<Rdyz, then the power grid is judged to be in normal operation and no processing is required; If the new energy disturbance score Rd is greater than or equal to the disturbance threshold Rdyz, that is, Rd ≥ Rdyz, the power grid is determined to be in an abnormal operating state, an automatic alarm is issued, the abnormal disturbance data is recorded, and the collaborative optimization module is triggered.

[0012] Preferably, the collaborative optimization module includes a reactive compensation calculation unit and an optimization unit; The reactive power compensation calculation unit is used to obtain the preliminary reactive power compensation amount Q of each inverter in the power grid when it works independently based on the power grid related data set S. Taking inverter i as an example, the preliminary reactive power compensation amount Q is obtained as follows: ; Where, 、 、 and Represents voltage deviation , the regression coefficients of humidity attenuation coefficient Sd, local cross interference factor Gr, and micro-arcing interference factor Dh; According to the initial reactive compensation amount Q of each inverter in the grid when working independently, the reactive conflict between the inverters is obtained. , among which, reactive power conflict The way to obtain it is: ; Where, and They represent the preliminary reactive compensation amounts of inverter j and inverter i respectively; Based on reactive power conflicts between inverters , construct the conflict matrix C, where the conflict matrix C is specifically expressed as: ; Where, the diagonal elements indicate no conflict, Indicates the degree of influence of reactive power adjustment of inverter 1 on inverter 2.

[0013] Preferably, the optimization unit is used to construct an objective function for the purpose of reducing reactive compensation conflicts. , where the objective function The specific manifestations are: ; Where, Represents the reactive compensation influence coefficient, represents the voltage deviation at inverter i, represents the error term, i.e., the difference between the voltage deviation and the voltage deviation compensated by the currently optimized reactive power compensation amount, N represents the total number of inverters, i={1, 2, 3, ..., N}, represents the reactive compensation amount Q that minimizes the objective function; Based on the constructed objective function , perform iterative optimization, the specific optimization process is as follows: Set the optimization step size and convergence threshold , and for each inverter, taking inverter i as an example, according to the reactive compensation conflict of inverter i , update the reactive compensation to obtain the updated reactive compensation of inverter i , where the reactive power compensation after the inverter i is updated is : ; Where, represents the initial reactive compensation amount of inverter i, represents the optimization step size

[0014] Reactive power compensation after inverter update , get the error term of each inverter And get the updated objective function value , and obtain the objective function difference after each inverter update , where the objective function difference The way to obtain it is: ; Where, Indicates the The objective function value after the reactive power compensation is updated, Indicates the The objective function value after the reactive power compensation is updated; The objective function difference and convergence threshold Compare and judge whether the iterative optimization has converged. The specific evaluation process is as follows: If the objective function difference Less than the convergence threshold , it is judged that the iterative optimization has converged, the optimization is stopped, and the reactive compensation of the inverter at this time is output. , and construct a reactive power compensation scheme, where the reactive power compensation scheme includes reactive power compensation for each inverter ; If the objective function difference Greater than or equal to the convergence threshold , it is determined that the iterative optimization has not converged, and the iterative optimization is continued until the objective function difference Less than the convergence threshold Or when the maximum number of iterations is reached, the iterative optimization is stopped.

[0015] Preferably, the adaptive reactive power compensation control module includes a short-time reactive power prediction unit and a micro-arc interference compensation optimization unit; The short-term reactive power prediction unit is used to build a short-term prediction model using the ARIMA algorithm and use historical power grid data to train the short-term prediction model to obtain the power grid frequency at the future time point t+k. , and calculate the frequency offset at the future time point t+k , where the frequency offset at the future time point t+k is The way to obtain it is: ; Where, Indicates the grid frequency at the current time point t; Based on the frequency offset at the future time point t+k and micro-arc interference factor Dh to obtain short-time optimized reactive compensation value , where the short-time optimized reactive compensation value is The way to obtain it is: ; Where, represents the reactive power compensation of inverter i at the current time point t, Indicates a future time point The frequency offset when represents the voltage of inverter i, Indicates the rated voltage of the grid. represents the nonlinear compensation adjustment coefficient, Indicates the sum of voltage deviations of all inverters.

[0016] Preferably, the micro-arc interference compensation optimization unit is used to optimize the reactive compensation value according to the short-term , obtain the reactive compensation error value , where the reactive compensation error value is The way to obtain it is: ; And according to the reactive compensation error value , get the final reactive compensation value : ; Based on the final reactive power compensation value of all inverters obtained , generate energy scheduling execution plan.

[0017] Preferably, the virtual power plant scheduling evaluation module is used to execute energy scheduling operations according to the energy scheduling execution plan, and continuously monitor the grid-related data during the scheduling process, perform summary calculations, and obtain the grid stability score PF, wherein the grid stability score PF is obtained in the following manner: ; Where, Indicates a future time point The grid frequency deviation at Indicates the inverter i at a future time point The voltage deviation when Indicates that the inverter with the largest voltage deviation among N inverters is selected. ∈{ , }; The grid stability threshold WD is preset, and the grid stability score PF and the grid stability threshold WD are compared and analyzed to evaluate the grid stability. The specific evaluation contents are as follows: If the grid stability score PF is less than the grid stability threshold WD, the grid stability is determined to be normal and the reactive power compensation adjustment is completed. At this time, the reactive power compensation related data is stored and the future grid stability is continuously monitored; If the grid stability score PF is greater than or equal to the grid stability threshold WD, the grid stability is judged to be in an abnormal state and there are frequency and voltage fluctuations in the grid. At this time, data is collected and fed back to the collaborative optimization module to regenerate the reactive power compensation plan.

[0018] Preferably, a distributed energy intelligent optimization scheduling method based on a virtual power plant comprises the following steps: Step 1: Collect and pre-process grid-related data to construct a grid-related data set S; Step 2: Based on the grid-related data set S, feature extraction is performed to obtain the humidity attenuation coefficient Sd, the local cross-interference factor Gr, and the micro-arcing interference factor Dh, and the new energy disturbance score Rd to evaluate the grid operation status; Step 3: Calculate the reactive power compensation amount of each inverter and generate a reactive power compensation plan for the power grid; Step 4: Optimize the scheme based on the reactive power compensation scheme of the power grid to obtain the energy dispatch execution plan; Step 5: Based on the energy dispatch execution plan, perform virtual power plant energy dispatch and conduct grid stability assessment.

[0019] The present invention provides a distributed energy intelligent optimization scheduling system and method based on a virtual power plant, which has the following beneficial effects: (1) Through the data acquisition and preprocessing module and the feature extraction and evaluation module, it can monitor in real time the impact of new energy (wind power, photovoltaic) on the power grid after grid connection, especially the voltage deviation, frequency drift and electromagnetic interference caused by the output fluctuation of new energy. Based on the humidity attenuation coefficient Sd, the local cross interference factor Gr and the micro arc interference factor The renewable energy disturbance score Rd calculated by Dh can predict the grid status in advance and trigger optimized scheduling when the grid is abnormal, thereby improving the system stability after the renewable energy is connected to the grid. Compared with traditional methods, this system can dynamically adapt to the power fluctuations of renewable energy, reduce the reactive power regulation lag problem, and effectively suppress the abnormal fluctuations of grid frequency and voltage caused by the access of renewable energy.

[0020] (2) The reactive compensation calculation unit and optimization unit are used to calculate the reactive compensation amount Q of each inverter, and the reactive compensation conflict between different inverters is evaluated through the conflict matrix C. The objective function is constructed for iterative optimization. Compared with the traditional single inverter control strategy, the optimization calculation of this system can ensure that each inverter reduces mutual interference when compensating reactive power, making reactive scheduling more coordinated. By optimizing the objective function and iterative calculation of the step size, this system can effectively reduce the reactive compensation conflict of the inverters within the power grid, improve the overall coordination of reactive power compensation, and improve the power quality and operation efficiency of the power grid.

[0021] (3) Adopting the adaptive reactive power compensation control module, further combining the short-time reactive power prediction unit and the micro-arc interference compensation optimization unit, using the short-time prediction model ARIMA to predict future grid frequency changes, and combining the micro-arc interference factor Dh to calculate the optimized compensation value , ensuring the accuracy and robustness of reactive compensation, the final compensation amount After correcting the micro-arc interference error, the execution accuracy of the compensation scheme is improved. Compared with the traditional fixed compensation strategy, this system can dynamically adapt to the reactive power demand of the power grid under the new energy environment, improve the reactive power compensation control accuracy, reduce the compensation error, improve the compensation execution effect, and make the power grid operation more stable and reliable. BRIEF DESCRIPTION OF THE DRAWINGS

[0022] Figure 1This is a block diagram of a distributed energy intelligent optimization scheduling system based on a virtual power plant in the present invention.

[0023] Figure 2 This is a flow chart of a distributed energy intelligent optimization scheduling method based on a virtual power plant according to the present invention.

[0024] Figure 3 This is a flow chart of the feature extraction and evaluation module of a distributed energy intelligent optimization scheduling system based on a virtual power plant in the present invention.

[0025] Figure 4 This is a line graph of the grid voltage fluctuation and reactive power compensation change trend of the present invention. DETAILED DESCRIPTION

[0026] The following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative efforts are within the scope of protection of the present invention.

[0027] Example 1 See also Figure 1 , the present invention provides a distributed energy intelligent optimization and scheduling system based on a virtual power plant, including a data acquisition and preprocessing module, a feature extraction and evaluation module, a collaborative optimization module, an adaptive reactive power compensation control module and a virtual power plant scheduling evaluation module; The data acquisition and preprocessing module is used to collect power grid related data, perform preprocessing, and construct a power grid related data set S; The feature extraction and evaluation module is used to extract features based on the grid-related data set S, obtain the humidity attenuation coefficient Sd, the local cross interference factor Gr, and the micro-arc interference factor Dh, and obtain the new energy disturbance score Rd to evaluate the grid operation status; The collaborative optimization module is used to calculate the reactive compensation amount of each inverter and generate a reactive compensation plan for the power grid; The adaptive reactive power compensation control module is used to optimize the reactive power compensation scheme according to the power grid to obtain the energy scheduling execution plan; The virtual power plant dispatch evaluation module is used to perform virtual power plant energy dispatch and grid stability evaluation based on the energy dispatch execution plan.

[0028] In the embodiment, the data acquisition and preprocessing module can monitor the operating status of the power grid in real time, including renewable energy power generation data, environmental parameters and grid status data, and improve the reliability and accuracy of the data through data cleaning and standardization. Subsequently, the feature extraction and evaluation module uses the renewable energy disturbance score Rd to evaluate the grid status, ensuring that potential grid instability factors caused by humidity attenuation, electromagnetic interference and micro-arcing phenomena can be identified, thereby improving the prediction accuracy. The collaborative optimization module optimizes reactive compensation by constructing a conflict matrix C, reduces reactive compensation conflicts between inverters, improves the coordination of the compensation strategy, and makes reactive power optimization more accurate. Combined with the adaptive reactive compensation control module, the compensation amount is further optimized based on short-term frequency prediction and micro-arcing interference correction, so that the compensation strategy dynamically adapts to the fluctuation of renewable energy output and improves compensation accuracy. Finally, the virtual power plant scheduling evaluation module executes the optimized reactive compensation scheme and monitors the scheduling effect through grid stability evaluation to ensure the stability and safety of system operation. Compared with traditional methods, this system improves the adaptability of renewable energy grid connection through full-process optimization, effectively reduces voltage fluctuations and frequency drift, enhances the intelligence level of grid scheduling, and improves the energy management efficiency and stability of the virtual power plant.

[0029] Example 2 Please refer to Figure 1 、 Figure 3 and Figure 4 ,Specifically: the data acquisition and preprocessing module includes a data acquisition unit and a ,preprocessing unit; The data acquisition unit is used to determine the power grid topology of the target area and select several power grid nodes. Intelligent sensor groups are deployed at these nodes to collect power grid-related data in real time. Grid nodes include substations, grid connection points, distribution rooms, wind turbines, and photovoltaic power plants. The intelligent sensor groups include synchronized phasor measurement devices, temperature sensors, and humidity sensors. Power grid-related data includes power grid operation data, new energy data, and environmental data. Grid operation data including voltage V, frequency f, inverter current I and electromagnetic interference intensity B are obtained by using synchronized phasor measurement devices; New energy data includes wind turbine output power and photovoltaic output power , obtained by using a synchronized phasor measurement device; Environmental data include temperature T and humidity RH, which are obtained by using temperature sensors and humidity sensors respectively; The preprocessing unit is used to perform noise filtering, outlier removal and data standardization on the collected power grid related data, and to construct a power grid related data set S based on the preprocessed power grid related data.

[0030] In the embodiment, through the coordinated work of the data acquisition unit and the pre-processing unit, high-precision monitoring of the grid operation status and data optimization processing can be achieved to ensure the accuracy and stability of subsequent optimization scheduling calculations. The data acquisition unit can flexibly determine the grid topology of the target area and deploy intelligent sensor groups at key grid nodes such as substations, grid connection points, distribution rooms, wind turbines and photovoltaic power stations to achieve all-round perception of grid operation data, new energy data and environmental data. The voltage V, frequency f, inverter current I and electromagnetic interference intensity B are collected through a synchronized phasor measurement unit (PMU), and combined with the output power of the wind turbine , photovoltaic output power , temperature T and humidity RH, the system can capture the dynamic changes of the power grid brought about by the grid connection of new energy. The preprocessing unit further filters the noise, removes outliers and standardizes the collected data to eliminate sensor measurement errors, improve the reliability and consistency of the data, and finally constructs the power grid related data set S. Compared with the traditional single data monitoring method, this module not only improves the comprehensiveness and accuracy of power grid status monitoring, but also ensures the accuracy of subsequent feature extraction, optimization calculation and reactive power compensation strategy, providing more reliable data support for intelligent scheduling, thereby improving the new energy absorption capacity, optimizing the reactive power scheduling of the power grid, and improving the stability and safety of power grid operation.

[0031] Example 3 Please refer to Figure 1 and Figure 3 ,Specifically: the feature extraction and evaluation module includes a ,feature extraction unit and a preliminary evaluation unit; The feature extraction unit is used to extract features based on the grid-related data set S to obtain the humidity attenuation coefficient Sd, the local cross interference factor Gr, and the micro-arcing interference factor Dh. The humidity attenuation coefficient Sd is obtained as follows: ; Where, Indicates the humidity of the target area at the current time point t, Indicates the temperature of the target area at the current time point t, Indicates the historical maximum humidity of the target area, and Represent the historical minimum and maximum temperatures of the target area, is the humidity normalization term, which is used to express the proportion of humidity relative to the maximum humidity in the past. e represents the base of the natural logarithm. represents an exponential term used to describe the moderating effect of temperature on humidity; The integration of renewable energy into the grid can lead to frequency shifts, local voltage imbalances, and power dispatch delays. The humidity attenuation coefficient (Sd) of high-voltage lines directly affects the actual transmission capacity of transmission lines, thereby affecting the voltage stability of the grid. If the impact of humidity on transmission lines is not considered, some renewable energy power may not be effectively delivered, affecting power balance, causing local voltage increases or decreases, and further affecting grid frequency. The humidity attenuation coefficient quantifies the impact of humidity on transmission lines, making reactive power compensation strategies more precise, enabling the system to adapt to different humidity environments, improving the reactive power dispatch capability of the grid, and reducing the stability risks brought about by the integration of renewable energy. The local cross interference factor Gr is obtained as follows: ; Where, Indicates the electromagnetic interference intensity of the target area at the current time point t, and Respectively represent the historical minimum and maximum electromagnetic interference intensity of the target area, represents the output power of wind turbines in the target area at the current time point t, represents the photovoltaic output power of the target area at the current time point t, and They represent the historical maximum wind turbine output power and maximum photovoltaic output power in the target area, Indicates the normalization of electromagnetic interference effects. Indicates that the difference between wind power and photovoltaic output is normalized.

[0032] Distributed renewable energy devices, including wind turbines and photovoltaic power plants, are connected to the smart grid through inverters. The switching actions and pulse signals of the inverters will form electromagnetic fields in the surrounding space. These electromagnetic fields may interfere with adjacent renewable energy devices, affecting the flow of reactive power in the grid and causing voltage fluctuations in local areas. Traditional reactive power compensation methods are based on voltage / reactive power curves and do not take into account the mutual influence between renewable energy devices. This may cause inverter scheduling and measurement conflicts, thereby triggering oscillations. The local cross-interference factor Gr dynamically adjusts the reactive compensation amount by considering the difference in wind power and photovoltaic output. When the electromagnetic interference intensity is high, reactive compensation is automatically increased to prevent local voltage instability. It quantifies the difference in renewable energy output and electromagnetic interference, allowing the reactive compensation system to dynamically adapt to the uncertainty of renewable energy grid connection, enabling the reactive compensation system to respond to renewable energy fluctuations in a timely manner, improving grid stability, and reducing voltage shocks caused by renewable energy access. The micro-arc interference factor Dh is obtained as follows: ; Where, represents the inverter current value of the target area at the current time point t, Indicates the inverter's power consumption per unit time The current variation amplitude, and They represent the historical maximum current value and minimum current value of the inverter in the target area respectively.

[0033] In an environment with high penetration of renewable energy and grid connection, the inverter's reactive power regulation is key to stabilizing grid frequency and voltage. However, when micro-arc discharges occur within the inverter, its current fluctuates unstably. This can cause errors in the reactive compensation calculated by the inverter due to arc interference, weakening the grid's voltage regulation and even causing further instability. Furthermore, if the system continues to perform reactive compensation according to normal strategies even when experiencing high micro-arc interference, it can further destabilize the grid, leading to scheduling lags or secondary adjustments. When the micro-arc interference factor Dh increases, the system can sense current instability and prevent voltage anomalies caused by miscompensation. By combining the micro-arc interference factor Dh with the reactive compensation calculation, the system can make grid scheduling more accurate and reduce secondary adjustments caused by miscompensation.

[0034] The preliminary evaluation unit is used to perform summary calculations based on the humidity attenuation coefficient Sd, the local cross interference factor Gr, and the micro-arcing interference factor Dh to obtain the new energy disturbance score Rd. The new energy disturbance score Rd is obtained as follows: ; The humidity attenuation coefficient Sd, the local cross-interference factor Gr, and the micro-arc interference factor Dh have different influence directions. The square sum calculation method can avoid the mutual offset of different disturbance factors, ensure the accuracy of the evaluation results, and highlight factors with larger values. When a factor is too large, the new energy disturbance score Rd will also increase rapidly, making it easier to detect abnormal conditions of new energy disturbances more quickly.

[0035] The disturbance threshold Rdyz is preset, and the new energy disturbance score Rd and the disturbance threshold Rdyz are compared and analyzed to evaluate the grid operation status. The specific evaluation contents are as follows: If the new energy disturbance score Rd is less than the disturbance threshold Rdyz, that is, Rd<Rdyz, then the power grid is judged to be in normal operation and no processing is required; If the new energy disturbance score Rd is greater than or equal to the disturbance threshold Rdyz, that is, Rd ≥ Rdyz, the power grid is determined to be in an abnormal operating state, an automatic alarm is issued, the abnormal disturbance data is recorded, and the collaborative optimization module is triggered.

[0036] In an embodiment, through the feature extraction unit, the system can calculate the humidity attenuation coefficient Sd, the local cross-interference factor Gr, and the micro-arc interference factor Dh based on the grid-related data set S, and quantify the key factors affecting the stability of new energy grid connection. Among them, the humidity attenuation coefficient Sd reflects the impact of humidity and temperature on the performance of grid equipment, the local cross-interference factor Gr measures the coupling relationship between electromagnetic interference and new energy output, and the micro-arc interference factor Dh is used to evaluate the volatility of the inverter output. Subsequently, the preliminary evaluation unit calculates the new energy disturbance score Rd based on these three factors, and compares it with the preset disturbance threshold Rdyz to accurately judge the operating status of the grid. When the grid operation is stable, the system can maintain the existing scheduling strategy; when the disturbance score exceeds the safety threshold, the system will automatically alarm, record abnormal data, and trigger the collaborative optimization module to adjust the reactive compensation. Compared with the traditional scheduling method that relies on fixed experience values, this module can dynamically identify the grid status, predict in advance the instability factors that may be caused by the grid connection of new energy, reduce scheduling delays, and improve the safety and reliability of grid operation.

[0037] Example 4 Please refer to Figure 1 and Figure 4 ,Specifically: the collaborative optimization module includes a reactive compensation ,computation unit and an optimization unit; The reactive power compensation calculation unit is used to obtain the preliminary reactive power compensation amount Q of each inverter in the power grid when it works independently based on the power grid related data set S. Taking inverter i as an example, the preliminary reactive power compensation amount Q is obtained as follows: ; Where, 、 、 and Represents voltage deviation , the regression coefficient of humidity attenuation coefficient Sd, local cross interference factor Gr and micro arc interference factor Dh, where the regression coefficient 、 、 and The multivariate linear regression method is used to obtain the historical power operation data; When the voltage decreases, the reactive compensation amount needs to be increased, and when the voltage increases, the reactive compensation amount needs to be reduced. The two are inversely proportional. The voltage deviation Use a minus sign before; High humidity will affect the insulation performance of the power grid line, increase line loss, and thus increase reactive power loss. Additional reactive power compensation is required to maintain voltage stability. A negative sign is used before the humidity attenuation coefficient Sd. Cross-interference causes uneven reactive power distribution, requiring additional compensation. The larger the local cross-interference factor Gr, the greater the reactive power compensation requirement, and a negative sign is used; Micro-arcing interference will cause the local voltage to rise rather than fall. When the micro-arcing interference increases, it leads to a reduction in reactive power compensation demand, and a positive sign is used.

[0038] According to the initial reactive compensation amount Q of each inverter in the grid when working independently, the reactive conflict between the inverters is obtained. , among which, reactive power conflict The way to obtain it is: ; Where, and They represent the preliminary reactive compensation amounts of inverter j and inverter i respectively; Based on reactive power conflicts between inverters , construct the conflict matrix C, where the conflict matrix C is specifically expressed as: ; In the formula, the diagonal elements indicate no conflict. The reactive power regulation of the inverter itself will not cause conflict with itself, so the diagonal elements are set to 0. Indicates the degree of influence of reactive power adjustment of inverter 1 on inverter 2. Indicates the degree of influence of reactive power adjustment of inverter 1 on inverter 2. Indicates the degree of influence of reactive power adjustment of inverter 1 on inverter 3, Indicates the degree of influence of reactive power adjustment of inverter 2 on inverter 1. Indicates the degree of influence of reactive power adjustment of inverter 2 on inverter 3, Indicates the degree of influence of reactive power adjustment of inverter 3 on inverter 1, Indicates the degree of influence of reactive power adjustment of inverter 3 on inverter 2.

[0039] The optimization unit is used to construct the objective function with the purpose of reducing reactive compensation conflicts. , where the objective function The specific manifestations are: ; Where, Represents the reactive compensation influence coefficient, represents the voltage deviation at inverter i, represents the error term, i.e., the difference between the voltage deviation and the voltage deviation compensated by the currently optimized reactive power compensation amount, N represents the total number of inverters, i={1, 2, 3, ..., N}, represents the reactive compensation amount Q that minimizes the objective function; Reactive compensation influence coefficient Obtained by linear data regression modeling using historical data; Based on the constructed objective function , perform iterative optimization, the specific optimization process is as follows: Set the optimization step size and convergence threshold , and for each inverter, taking inverter i as an example, according to the reactive compensation conflict of inverter i , update the reactive compensation to obtain the updated reactive compensation of inverter i , where the reactive power compensation after the inverter i is updated is : ; Where, represents the initial reactive compensation amount of inverter i, represents the optimization step size

[0040] Reactive power compensation after inverter update , get the error term of each inverter , and obtain the updated objective function value , and obtain the objective function difference after each inverter update , where the objective function difference The way to obtain it is: ; Where, Indicates the The objective function value after the reactive power compensation is updated, Indicates the The objective function value after the reactive power compensation is updated; The objective function difference and convergence threshold Compare and judge whether the iterative optimization has converged. The specific evaluation process is as follows: If the objective function difference Less than the convergence threshold , it is judged that the iterative optimization has converged, the optimization is stopped, and the reactive compensation of the inverter at this time is output. , and construct a reactive power compensation scheme, where the reactive power compensation scheme includes reactive power compensation for each inverter ; If the objective function difference Greater than or equal to the convergence threshold , it is determined that the iterative optimization has not converged, and the iterative optimization is continued until the objective function difference Less than the convergence threshold Or when the maximum number of iterations is reached, the iterative optimization is stopped. The maximum number of iterations is set by the customer based on actual conditions.

[0041] The following are some examples: Assume that three inverters are connected to the grid and numbered 1, 2, and 3 respectively. The initial reactive power compensation values ​​of the inverters are obtained as shown in Table 1 below:

[0042] Table 1 Calculate the initial error terms separately and get the results as shown in Table 2:

[0043] Table 2 According to Table 2 above, get the initial objective function value =0.0014, set the optimization step size When the value is 0.1, the reactive compensation is updated and the updated reactive compensation value of each inverter is obtained: =3.998Var, =3.501Var, =4.997Var; And calculate the new error term, and get the result 4 as shown in Table 3:

[0044] Table 3 Setting the convergence threshold =0.001, get the new objective function value =0.0011, and calculate the objective function difference =0.000266, at this time the objective function difference =0.000266 is less than the convergence threshold , terminate the optimization process, output the reactive compensation amount of the inverter at this time, and build a reactive compensation scheme.

[0045] In the embodiment, the reactive compensation calculation unit and the optimization unit of the collaborative optimization module are used to coordinate reactive compensation strategies among multiple inverters to reduce compensation conflicts and improve grid operation stability. First, the reactive compensation calculation unit can calculate the reactive power according to the voltage deviation. , humidity attenuation coefficient Sd, local cross-interference factor Gr, and micro-arcing interference factor Dh to calculate the initial reactive compensation amount Q, ensuring that the compensation scheme accurately reflects the dynamic state of the power grid. Subsequently, the system calculates the degree of reactive compensation conflict between inverters based on the reactive conflict matrix C and optimizes the compensation strategy to achieve a more reasonable distribution of reactive compensation between different inverters. The objective function is constructed through optimization units. During the iterative process, the system continuously adjusts the compensation amount to reduce compensation conflicts and improve the overall compensation effect. During the optimization calculation process, the optimization step size β and convergence threshold ϵ are used to control the stability of the optimization process and ensure the feasibility of the final reactive compensation scheme. If the optimized reactive compensation adjustment meets the convergence conditions, a reactive compensation scheme is constructed to improve reactive compensation accuracy and system stability. If the optimization does not converge, the compensation amount is automatically adjusted and iterative optimization continues to ensure the reliability of power grid operation. Compared with traditional reactive compensation methods, this system can dynamically adapt to the fluctuations of renewable energy after grid integration, reduce reactive power conflicts caused by renewable energy power fluctuations, improve the coordination ability between inverters, make the power grid more stable and efficient, and ensure power quality and power supply reliability.

[0046] Example 5 Please refer to Figure 1 and Figure 4 ,Specifically: the adaptive reactive compensation control module includes a short-time reactive ,prediction unit and a micro-arc interference compensation optimization unit; The short-term reactive power prediction unit is used to build a short-term prediction model using the ARIMA algorithm and use historical power grid data to train the short-term prediction model to obtain the power grid frequency at the future time point t+k. , and calculate the frequency offset at the future time point t+k , where the frequency offset at the future time point t+k is The way to obtain it is: ; Where, Indicates the grid frequency at the current time point t; Based on the frequency offset at the future time point t+k and micro-arc interference factor Dh to obtain short-time optimized reactive compensation value , where the short-time optimized reactive compensation value is The way to obtain it is: ; Where, represents the reactive power compensation of inverter i at the current time point t, Indicates a future time point The frequency offset when represents the voltage of inverter i, Indicates the rated voltage of the grid. represents the nonlinear compensation adjustment coefficient, represents the sum of voltage deviations of all inverters, It represents the predicted frequency drift correction term, combined with the micro-arc interference factor Dh, to correct the reactive compensation error caused by micro-arc interference in short-term prediction, and optimize the reactive compensation value in a short time. It solves the sensitivity problems of new energy grid-connected fluctuations, reactive power compensation lag and micro-arcing affecting miscompensation, and improves the stability of grid operation.

[0047] Nonlinear compensation adjustment coefficient Used to amplify and reduce the impact of the total deviation on reactive power compensation. If the total voltage deviation of all inverters is Large, nonlinear compensation adjustment coefficient It will amplify the reactive compensation adjustment strength, so that the inverter can provide more reactive power to maintain the grid voltage stability. Among them, the nonlinear compensation adjustment coefficient is a sum of voltage deviations The dynamic nonlinear regulation function, the regulation algorithm is specifically expressed as follows: ; Where, Indicates the total voltage deviation , Indicates the critical value of the total voltage deviation, Indicates the low deviation segment amplification factor, Indicates the high deviation segment amplification factor, represents the exponential regulation rate coefficient, represents the logarithmic adjustment scale factor, e represents the base of the natural logarithm, and ln represents the natural logarithm function; Voltage deviation sum critical value The low deviation amplification factor is obtained by the customer through superposition calculation based on the inverter data according to the "National Standard for Power Quality of Distribution Networks". , exponential adjustment rate coefficient , high deviation section amplification factor and logarithmic scaling factors Obtained by fitting regression using historical operating data; Nonlinear compensation adjustment coefficient The voltage deviation sum Drive, using piecewise function for adjustment, when 0≤total voltage deviation ≤ voltage deviation total critical value When the nonlinear compensation adjustment coefficient The exponential function is used to slowly increase the voltage to avoid excessive amplification of small deviations. > Voltage deviation total critical value When the nonlinear compensation adjustment coefficient Rapid enhancement of the natural logarithmic function enables fast compensation response under large deviations.

[0048] The micro-arc interference compensation optimization unit is used to optimize the reactive compensation value based on short-term , obtain the reactive compensation error value , where the reactive compensation error value is The way to obtain it is: ; And according to the reactive compensation error value , get the final reactive compensation value : ; Based on the final reactive power compensation value of all inverters obtained , generate energy scheduling execution plan.

[0049] In the embodiment, a more accurate and dynamic reactive compensation optimization is achieved through the short-time reactive power prediction unit and the micro-arc interference compensation optimization unit, ensuring the stability of the power grid in an environment with high penetration of new energy. First, the short-time reactive power prediction unit adopts the ARIMA short-time prediction algorithm, uses historical power grid data to train the prediction model, and obtains the power grid frequency at the future time point t+k. , and calculate the frequency offset at the future time point t+k , thereby sensing the short-term fluctuation trend of the power grid in advance, combined with the predicted frequency offset And the micro-arc interference factor Dh, the system further calculates the short-time optimized reactive compensation value , ensuring that the compensation strategy can adapt to short-term dynamic changes in the power grid and reduce the compensation error caused by hysteresis adjustment. Secondly, the micro-arc interference compensation optimization unit further corrects the reactive compensation error and makes compensation adjustments for the error caused by micro-arc interference to improve the compensation accuracy. The final reactive compensation value is calculated through optimization. , so that the compensation scheme not only takes into account short-term dynamic frequency fluctuations, but also effectively combats micro-arc interference, improves the execution accuracy of reactive compensation and grid stability. Compared with traditional fixed compensation methods, the system can dynamically and adaptively adjust the compensation strategy, improve the grid frequency and voltage stability under new energy environments, reduce system instability problems caused by reactive power regulation lag, optimize grid power quality, and improve scheduling reliability and compensation accuracy.

[0050] Example 6 Please refer to Figure 1Specifically, the virtual power plant dispatch evaluation module is used to execute energy dispatch operations according to the energy dispatch execution plan, and continuously monitor the grid-related data during the dispatch process, perform summary calculations, and obtain the grid stability score PF. The grid stability score PF is obtained in the following way: ; Where, Indicates a future time point The grid frequency deviation at Indicates the inverter i at a future time point The voltage deviation at Indicates that the inverter with the largest voltage deviation among N inverters is selected. ∈{ , }; The grid stability threshold WD is preset, and the grid stability score PF and the grid stability threshold WD are compared and analyzed to evaluate the grid stability. The specific evaluation contents are as follows: If the grid stability score PF is less than the grid stability threshold WD, the grid stability is determined to be normal and the reactive power compensation adjustment is completed. At this time, the reactive power compensation related data is stored and the future grid stability is continuously monitored; If the grid stability score PF is greater than or equal to the grid stability threshold WD, the grid stability is judged to be in an abnormal state and there are frequency and voltage fluctuations in the grid. At this time, data is collected and fed back to the collaborative optimization module to regenerate the reactive power compensation plan.

[0051] In the embodiment, the precise control of grid stability is achieved by continuously monitoring and dynamically evaluating the executed reactive power compensation scheme. The module adopts the grid stability score PF calculation formula, which comprehensively considers the grid frequency deviation at the future time point. and inverter voltage deviation , ensuring that the actual implementation effect of the reactive compensation scheme meets the operation requirements of the power grid. Compared with the traditional static compensation method, this system can not only analyze the scheduling effect of reactive compensation in real time, but also continuously track the dynamic changes of the power grid, making reactive regulation more accurate and reliable. Through this dynamic evaluation and adjustment mechanism, the module can improve the intelligence level of power grid scheduling, effectively reduce reactive compensation errors in the new energy grid-connected environment, improve the frequency and voltage stability of the power grid, and ultimately enhance the reliability and safety of the power grid.

[0052] Example 7 Please refer to Figure 2 Specifically: A distributed energy intelligent optimization scheduling method based on virtual power plant, including the following steps: Step 1: Collect and pre-process grid-related data to construct a grid-related data set S; Step 2: Based on the grid-related data set S, feature extraction is performed to obtain the humidity attenuation coefficient Sd, the local cross-interference factor Gr, and the micro-arcing interference factor Dh, and the new energy disturbance score Rd to evaluate the grid operation status; Step 3: Calculate the reactive power compensation amount of each inverter and generate a reactive power compensation plan for the power grid; Step 4: Optimize the scheme based on the reactive power compensation scheme of the power grid to obtain the energy dispatch execution plan; Step 5: Based on the energy dispatch execution plan, perform virtual power plant energy dispatch and conduct grid stability assessment.

[0053] In the embodiment, the grid operation data, renewable energy output and environmental factors are collected and preprocessed in real time to ensure the accuracy of the data. Subsequently, the renewable energy disturbance characteristics are extracted based on the humidity attenuation coefficient Sd, the local cross-interference factor Gr and the micro-arc interference factor Dh, the grid operation status is evaluated, and the impact of renewable energy grid connection on the grid is perceived in advance. In the reactive compensation calculation stage, the conflict matrix C and the iterative optimization algorithm are used to improve the coordination of the compensation strategy and reduce the reactive compensation conflict between inverters. Through adaptive compensation control, short-term prediction and interference correction are used to improve the accuracy of reactive scheduling. Finally, the grid stability is monitored in real time and the energy scheduling strategy is optimized, thereby improving the safety of renewable energy grid connection and the stability of grid operation, and enhancing the scheduling flexibility and economy of the system.

[0054] While embodiments of the present invention have been shown and described, it will be appreciated by those skilled in the art that various changes, modifications, substitutions, and variations may be made to these embodiments without departing from the principles and spirit of the invention, and that the scope of the invention is defined by the appended claims and their equivalents.

Claims

1. A distributed energy intelligent optimization and scheduling system based on a virtual power plant, characterized by: It includes data acquisition and preprocessing module, feature extraction and evaluation module, collaborative optimization module, adaptive reactive power compensation control module and virtual power plant scheduling evaluation module; The data acquisition and preprocessing module is used to collect power grid related data, perform preprocessing, and construct a power grid related data set S; The feature extraction and evaluation module is used to extract features based on the grid-related data set S, obtain the humidity attenuation coefficient Sd, the local cross interference factor Gr, and the micro-arc interference factor Dh, and obtain the new energy disturbance score Rd to evaluate the grid operation status; The collaborative optimization module is used to calculate the reactive compensation amount of each inverter and generate a reactive compensation plan for the power grid; The adaptive reactive power compensation control module is used to optimize the reactive power compensation scheme according to the power grid to obtain the energy scheduling execution plan; The virtual power plant dispatch evaluation module is used to perform virtual power plant energy dispatch and grid stability evaluation based on the energy dispatch execution plan.

2. A distributed energy intelligent optimization and scheduling system based on a virtual power plant according to claim 1, characterized in that: The data acquisition and preprocessing module includes a data acquisition unit and a preprocessing unit; The data acquisition unit is used to determine the power grid topology of the target area and select several power grid nodes. Intelligent sensor groups are deployed at these nodes to collect power grid-related data in real time. Grid nodes include substations, grid connection points, distribution rooms, wind turbines, and photovoltaic power plants. The intelligent sensor groups include synchronized phasor measurement devices, temperature sensors, and humidity sensors. Power grid-related data includes power grid operation data, new energy data, and environmental data. Grid operation data includes voltage V, frequency f, inverter current I, and electromagnetic interference intensity B; New energy data includes wind turbine output power and photovoltaic output power ; Environmental data include temperature T and humidity RH; The preprocessing unit is used to perform noise filtering, outlier removal and data standardization on the collected power grid related data, and construct a power grid related data set S based on the preprocessed power grid related data.

3. A distributed energy intelligent optimization and scheduling system based on a virtual power plant according to claim 2, characterized in that: The feature extraction and evaluation module includes a feature extraction unit and a preliminary evaluation unit; The feature extraction unit is used to extract features based on the grid-related data set S to obtain the humidity attenuation coefficient Sd, the local cross interference factor Gr, and the micro-arcing interference factor Dh. The humidity attenuation coefficient Sd is obtained as follows: ； Where, Indicates the humidity of the target area at the current time point t, Indicates the temperature of the target area at the current time point t, Indicates the historical maximum humidity of the target area, and They represent the historical minimum and maximum temperatures of the target area, respectively, and e represents the base of the natural logarithm. represents the exponential term; The local cross interference factor Gr is obtained as follows: ； Where, Indicates the electromagnetic interference intensity of the target area at the current time point t, and Respectively represent the historical minimum and maximum electromagnetic interference intensity of the target area, represents the output power of wind turbines in the target area at the current time point t, represents the photovoltaic output power of the target area at the current time point t, and They represent the historical maximum wind turbine output power and maximum photovoltaic output power in the target area respectively; The micro-arc interference factor Dh is obtained as follows: ； Where, represents the inverter current value of the target area at the current time point t, Indicates the inverter's power consumption per unit time The current variation amplitude, and They represent the historical maximum current value and minimum current value of the inverter in the target area respectively.

4. A distributed energy intelligent optimization and scheduling system based on a virtual power plant according to claim 3, characterized in that: The preliminary evaluation unit is used to perform summary calculations based on the humidity attenuation coefficient Sd, the local cross interference factor Gr, and the micro-arcing interference factor Dh to obtain the new energy disturbance score Rd. The new energy disturbance score Rd is obtained as follows: ； The disturbance threshold Rdyz is preset, and the new energy disturbance score Rd and the disturbance threshold Rdyz are compared and analyzed to evaluate the grid operation status. The specific evaluation contents are as follows: If the new energy disturbance score Rd is less than the disturbance threshold Rdyz, that is, Rd<Rdyz, then the power grid is judged to be in normal operation and no processing is required; If the new energy disturbance score Rd is greater than or equal to the disturbance threshold Rdyz, that is, Rd ≥ Rdyz, the power grid is determined to be in an abnormal operating state, an automatic alarm is issued, the abnormal disturbance data is recorded, and the collaborative optimization module is triggered.

5. A distributed energy intelligent optimization and dispatching system based on a virtual power plant according to claim 4, characterized in that: The collaborative optimization module includes a reactive compensation calculation unit and an optimization unit; The reactive power compensation calculation unit is used to obtain the preliminary reactive power compensation amount Q of each inverter in the power grid when it works independently based on the power grid related data set S. Taking inverter i as an example, the preliminary reactive power compensation amount Q is obtained as follows: ； Where, 、 、 and Represents voltage deviation , the regression coefficients of humidity attenuation coefficient Sd, local cross interference factor Gr, and micro-arcing interference factor Dh; According to the initial reactive compensation amount Q of each inverter in the grid when working independently, the reactive conflict between the inverters is obtained. , among which, reactive power conflict The way to obtain it is: ； Where, and They represent the preliminary reactive compensation amounts of inverter j and inverter i respectively; Based on reactive power conflicts between inverters , construct the conflict matrix C, where the conflict matrix C is specifically expressed as: ； Where, the diagonal elements indicate no conflict, Indicates the degree of influence of reactive power adjustment of inverter 1 on inverter 2.

6. A distributed energy intelligent optimization and dispatching system based on a virtual power plant according to claim 5, characterized in that: The optimization unit is used to construct the objective function with the purpose of reducing reactive compensation conflicts. , where the objective function The specific manifestations are: ； Where, Represents the reactive compensation influence coefficient, represents the voltage deviation at inverter i, represents the error term, i.e., the difference between the voltage deviation and the voltage deviation compensated by the currently optimized reactive power compensation amount, N represents the total number of inverters, i={1, 2, 3, ..., N}, represents the reactive compensation amount Q that minimizes the objective function; Based on the constructed objective function , perform iterative optimization, the specific optimization process is as follows: Set the optimization step size and convergence threshold , and for each inverter, taking inverter i as an example, according to the reactive compensation conflict of inverter i , update the reactive compensation to obtain the updated reactive compensation of inverter i , where the reactive power compensation after the inverter i is updated is The way to obtain it is: ； Where, represents the initial reactive compensation amount of inverter i, represents the optimization step size; Reactive power compensation after inverter update , get the error term of each inverter , and obtain the updated objective function value , and obtain the objective function difference after each inverter update , where the objective function difference The way to obtain it is: ； Where, Indicates the The objective function value after the reactive power compensation is updated, Indicates the The objective function value after the reactive power compensation is updated; The objective function difference and convergence threshold Compare and judge whether the iterative optimization has converged. The specific evaluation process is as follows: If the objective function difference Less than the convergence threshold , it is judged that the iterative optimization has converged, the optimization is stopped, and the reactive compensation of the inverter at this time is output. , and construct a reactive power compensation scheme, where the reactive power compensation scheme includes reactive power compensation for each inverter ; If the objective function difference Greater than or equal to the convergence threshold , it is determined that the iterative optimization has not converged, and the iterative optimization is continued until the objective function difference Less than the convergence threshold Or when the maximum number of iterations is reached, the iterative optimization is stopped.

7. A distributed energy intelligent optimization and dispatching system based on a virtual power plant according to claim 6, characterized in that: The adaptive reactive power compensation control module includes a short-time reactive power prediction unit and a micro-arc interference compensation optimization unit; The short-term reactive power prediction unit is used to build a short-term prediction model using the ARIMA algorithm and use historical power grid data to train the short-term prediction model to obtain the power grid frequency at the future time point t+k. , and calculate the frequency offset at the future time point t+k , where the frequency offset at the future time point t+k is The way to obtain it is: ； Where, Indicates the grid frequency at the current time point t; Based on the frequency offset at the future time point t+k and micro-arc interference factor Dh to obtain short-time optimized reactive compensation value , where the short-time optimized reactive compensation value is The way to obtain it is: ； Where, represents the reactive power compensation of inverter i at the current time point t, Indicates a future time point The frequency offset when represents the voltage of inverter i, Indicates the rated voltage of the power grid, represents the nonlinear compensation adjustment coefficient, Indicates the sum of voltage deviations of all inverters.

8. A distributed energy intelligent optimization and dispatching system based on a virtual power plant according to claim 7, characterized in that: The micro-arc interference compensation optimization unit is used to optimize the reactive compensation value based on short-term , obtain the reactive compensation error value , where the reactive compensation error value is The way to obtain it is: ； And according to the reactive compensation error value , get the final reactive compensation value : ； Based on the final reactive power compensation value of all inverters obtained , generate energy scheduling execution plan.

9. A distributed energy intelligent optimization and dispatching system based on a virtual power plant according to claim 8, characterized in that: The virtual power plant dispatch evaluation module is used to execute energy dispatch operations according to the energy dispatch execution plan, continuously monitor the grid-related data during the dispatch process, perform summary calculations, and obtain the grid stability score PF. The grid stability score PF is obtained in the following way: ； Where, Indicates a future time point The grid frequency deviation at Indicates the inverter i at a future time point The voltage deviation when Indicates that the inverter with the largest voltage deviation among N inverters is selected. ∈{ , }; The grid stability threshold WD is preset, and the grid stability score PF and the grid stability threshold WD are compared and analyzed to evaluate the grid stability. The specific evaluation contents are as follows: If the grid stability score PF is less than the grid stability threshold WD, the grid stability is determined to be normal and the reactive power compensation adjustment is completed. At this time, the reactive power compensation related data is stored and the future grid stability is continuously monitored; If the grid stability score PF is greater than or equal to the grid stability threshold WD, the grid stability is judged to be in an abnormal state and there are frequency and voltage fluctuations in the grid. At this time, data is collected and fed back to the collaborative optimization module to regenerate the reactive power compensation plan.

10. A distributed energy intelligent optimization scheduling method based on a virtual power plant, used to implement a distributed energy intelligent optimization scheduling system based on a virtual power plant according to any one of claims 1 to 9, characterized in that: The following steps are included: Step 1: Collect and pre-process grid-related data to construct a grid-related data set S; Step 2: Based on the grid-related data set S, feature extraction is performed to obtain the humidity attenuation coefficient Sd, the local cross interference factor Gr, and the micro-arcing interference factor Dh, and the new energy disturbance score Rd to evaluate the grid operation status; Step 3: Calculate the reactive power compensation amount of each inverter and generate a reactive power compensation plan for the power grid; Step 4: Optimize the scheme based on the reactive power compensation scheme of the power grid to obtain the energy dispatch execution plan; Step 5: Based on the energy dispatch execution plan, perform virtual power plant energy dispatch and conduct grid stability assessment.

Images