Virtual power plant operation regulation method and device adaptive to multiple distributed resources

Abstract

The application discloses a virtual power plant operation regulation method and device suitable for multiple distributed resources, relates to the technical field of virtual power plants, and comprises the following steps: constructing a power plant simulation space; predicting a predicted power consumption load sequence, a predicted power grid frequency sequence and a predicted renewable energy output ratio sequence in a preset time zone according to historical power grid operation data, and determining a regulation response speed demand; determining an adaptive resource combination strategy according to the regulation response speed demand, and optimizing an operation regulation scheme based on the adaptive resource combination strategy to maximize the regulation accuracy, and determining an optimal regulation scheme to perform operation regulation. The technical problem that the existing virtual power plant operation regulation method cannot accurately depict the operation constraints of multiple types of distributed resources in the power plant, the regulation strategy is disconnected with the actual equipment characteristics, and the regulation efficiency is not high is solved.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of virtual power plants, in particular to a virtual power plant operation regulation method and device suitable for multiple distributed resources. BACKGROUND

[0002] With the wide access of distributed energy, the role of virtual power plants in the power system is becoming more and more important. However, the existing virtual power plant operation regulation method is not accurate enough in analyzing and utilizing the grid operation data, and cannot accurately predict the changes of power load, grid frequency and renewable energy output ratio, so it is difficult to develop the optimal regulation scheme, and the regulation response speed and accuracy cannot meet the actual demand, resulting in low regulation efficiency, which easily causes the instability of power grid operation and increases the risk of power supply. SUMMARY

[0003] The embodiments of the present application provide a virtual power plant operation regulation method and device suitable for multiple distributed resources, which solves the technical problem that the existing virtual power plant operation regulation method is not accurate in describing the operation constraints of multiple types of distributed resources in the power plant, the regulation strategy is disconnected with the actual device characteristics, and the regulation efficiency is low.

[0004] The technical solutions of the present application to solve the above technical problems are as follows:

[0005] In a first aspect, the present application provides a virtual power plant operation regulation method suitable for multiple distributed resources, comprising:

[0006] constructing a power plant simulation space comprising a distributed resource layer, a monitoring and metering layer and an operation regulation layer in a three-dimensional simulation platform;

[0007] monitoring and obtaining historical grid operation data in a historical time zone through the monitoring and metering layer, predicting a predicted power load sequence, a predicted grid frequency sequence and a predicted renewable energy output ratio sequence in a preset time zone according to the historical grid operation data, and determining a regulation response speed requirement;

[0008] determining an adaptive resource combination strategy according to the regulation response speed requirement, and performing operation regulation scheme optimization in the power plant simulation space based on the adaptive resource combination strategy to maximize the regulation accuracy, and determining an optimal regulation scheme for operation regulation.

[0009] In a second aspect, the present application provides a virtual power plant operation regulation device suitable for multiple distributed resources, comprising:

[0010] a space construction module for constructing a power plant simulation space comprising a distributed resource layer, a monitoring and metering layer and an operation regulation layer in a three-dimensional simulation platform;

[0011] a sequence prediction module, configured to monitor and acquire historical power grid operation data in a historical time zone through the monitoring and metering layer, and predict a predicted power consumption load sequence, a predicted power grid frequency sequence and a predicted renewable energy output ratio sequence in a preset time zone according to the historical power grid operation data, and determine a control response speed demand;

[0012] a scheme execution module, configured to determine an adaptive resource combination strategy according to the control response speed demand, and perform operation control scheme optimization in the power plant simulation space based on the adaptive resource combination strategy to maximize the control accuracy, and determine an optimal control scheme for operation control.

[0013] The application provides one or more technical solutions, at least having the following technical effects or advantages:

[0014] The application provides a virtual power plant operation control method and device adaptive to multiple distributed resources. First, a power plant simulation space is constructed on a three-dimensional simulation platform to comprehensively and accurately simulate the actual operation environment of the virtual power plant, laying a solid foundation for subsequent control work. Second, historical power grid operation data is acquired through the monitoring and metering layer, and then a power consumption load predictor, a power grid frequency predictor and an energy output ratio predictor are constructed to more accurately predict the predicted power consumption load sequence, the predicted power grid frequency sequence and the predicted renewable energy output ratio sequence in the preset time zone. Finally, an adaptive resource combination strategy is determined according to the control response speed demand, and operation control scheme optimization is performed in the power plant simulation space, comprehensively considering multiple aspects such as instruction tracking error index, regulation performance index and power grid influence index, to ensure the comprehensiveness and accuracy of the control accuracy evaluation.

[0015] Through the above technical solution, the method effectively solves the problem that the existing virtual power plant operation control method cannot accurately depict the operation constraints of multiple types of distributed resources in the power plant and the control strategy is disconnected from the actual device characteristics, improves the virtual power plant operation control efficiency, enhances the stability of the power grid operation, reduces the risk of power supply, and provides protection for the efficient operation of the virtual power plant. BRIEF DESCRIPTION OF DRAWINGS

[0016] In order to more clearly illustrate the technical solutions in the embodiments of the application, the drawings needed in the embodiment description will be briefly introduced. Obviously, the drawings in the following description are only some embodiments of the application, and other drawings can be obtained by those skilled in the art without creative labor.

[0017] Figure 1 is a flowchart of a virtual power plant operation control method adaptive to multiple distributed resources provided by the embodiments of the application;

[0018] Figure 2 This is a schematic diagram of the structure of the virtual power plant operation and control device adapted to multiple distributed resources provided in the embodiments of this application.

[0019] The components represented by each number in the attached diagram are explained below:

[0020] Spatial construction module 11, sequence prediction module 12, scheme execution module 13. Detailed Implementation

[0021] This application provides a virtual power plant operation and control method and apparatus adapted to multiple distributed resources, which addresses the technical problem that existing virtual power plant operation and control methods do not accurately characterize the operational constraints of multiple types of distributed resources within the power plant, and the control strategies are out of sync with the actual equipment characteristics, resulting in low control efficiency.

[0022] The technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, and not all embodiments. Based on the embodiments of this application, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this application.

[0023] In the description of this application, the terms "first" and "second" are used for descriptive purposes only and should not be construed as indicating or implying relative importance or implicitly specifying the number of indicated technical features. Thus, a feature defined as "first" or "second" may explicitly or implicitly include one or more of the stated features. In the description of this application, "multiple" means two or more, unless otherwise explicitly specified.

[0024] In the description of this application, the term "for example" is used to mean "used as an example, illustration, or description." Any embodiment described as "for example" in this application is not necessarily to be construed as being more preferred or advantageous than other embodiments. The following description is provided to enable any person skilled in the art to make and use this application. Details are set forth in the following description for purposes of explanation. It should be understood that those skilled in the art will recognize that this application can be made without using these specific details. In other instances, well-known structures and processes will not be described in detail to avoid unnecessarily obscuring the description of this application. Therefore, this application is not intended to be limited to the embodiments shown, but is consistent with the broadest scope of the principles and features disclosed in this application.

[0025] Example 1, as Figure 1 As shown, this application provides a method for the operation and control of a virtual power plant adapted to multiple distributed resources, including:

[0026] S10: constructing a power plant simulation space containing a distributed resource layer, a monitoring and metering layer, and an operation and control layer in a three-dimensional simulation platform;

[0027] In the embodiment of the application, first, a power plant simulation space is constructed in a three-dimensional simulation platform. The power plant simulation space contains a distributed resource layer, a monitoring and metering layer, and an operation and control layer by simulating the environment in which an actual virtual power plant operates.

[0028] The distributed resource layer covers various types of distributed energy sources in the power plant, such as solar photovoltaic panels, wind turbine generators, small hydropower plants, etc. The monitoring and metering layer obtains the operation data of each energy source in the distributed resource layer in real time by simulating actual monitoring equipment and systems, including power generation power, voltage, current, etc., and can also monitor indicators such as power load and grid frequency in real time. The operation and control layer controls and reasonably allocates the energy sources in the distributed resource layer according to the data obtained by the monitoring and metering layer, so as to realize the efficient and stable operation of the virtual power plant.

[0029] By constructing the power plant simulation space, an analog environment is provided for subsequent prediction and control work, which helps to improve the efficiency and accuracy of the operation and control of the virtual power plant.

[0030] Specifically, step S10 in the method includes:

[0031] The distributed resource layer is constructed in the three-dimensional simulation platform, wherein the distributed resource layer includes a combined heat and power unit, an energy storage system, a wind power generation unit, and a photovoltaic power generation unit, and each type of resource includes a corresponding operation mathematical model, and the operation mathematical model at least contains a ramp rate, a response delay, an energy state, and an operation boundary constraint;

[0032] In the three-dimensional simulation platform, intelligent monitoring and metering devices deployed in each distributed resource are simulated to construct a monitoring and metering layer for collecting operation state data in real time;

[0033] The operation and control layer is constructed in the three-dimensional simulation platform, wherein the operation and control layer is used to receive the data of the monitoring and metering layer and perform subsequent operation and control;

[0034] The power plant simulation space is built based on the distributed resource layer, the monitoring and metering layer, and the operation and control layer.

[0035] In the embodiment of the application, first, a distributed resource layer is constructed in a three-dimensional simulation platform, including a combined heat and power unit, an energy storage system, a wind power generation unit, and a photovoltaic power generation unit, and a corresponding operation mathematical model is provided for each type of resource. The model describes the ramp rate, response delay, energy state, and operation boundary constraint parameters, and describes the operation characteristics of various types of distributed resources.

[0036] Secondly, intelligent monitoring and metering devices are simulated and deployed within the three-dimensional simulation platform. These devices are distributed at various distributed resources and can collect real-time operational state data, thereby constructing the monitoring and metering layer. The monitoring and metering layer can obtain operational data such as the output of distributed resources, load, state of charge of energy storage, and the heat-to-power ratio of combined heat and power, providing a basis for subsequent control decisions.

[0037] Thirdly, the operational control layer is constructed, which is responsible for receiving data from the monitoring and metering layer and controlling and dispatching energy from the distributed resource layer. The operational control layer formulates appropriate control strategies based on real-time demand of the power grid and operational state of distributed resources, to achieve efficient and stable operation of the virtual power plant.

[0038] Finally, based on the above-mentioned distributed resource layer, monitoring and metering layer, and operational control layer, a complete power plant simulation space is built. The power plant simulation space can highly restore the actual operational environment of the virtual power plant, providing an accurate and reliable simulation platform for subsequent prediction and control work. After the power plant simulation space is constructed, different control scenarios can be simulated using the simulation space to identify potential problems in advance and adjust control strategies in a timely manner. At the same time, through analysis of the simulation results, the configuration and operational parameters of distributed resources can be continuously optimized, further improving the overall performance of the virtual power plant.

[0039] S20: Monitor and obtain historical power grid operational data in the historical time zone through the monitoring and metering layer, and predict the predicted power consumption load sequence, predicted power grid frequency sequence, and predicted renewable energy output ratio sequence in the preset time zone based on the historical power grid operational data, and determine the control response speed requirement;

[0040] In the embodiments of the present application, after the power plant simulation space is constructed, the intelligent monitoring and metering devices of the monitoring and metering layer are used to continuously monitor and record the historical power grid operational data in the historical time zone. The operational data includes power consumption load, power grid frequency, and renewable energy output ratio, etc.

[0041] Firstly, data analysis and machine learning algorithms are used to predict the relevant sequences in the preset time zone. For the predicted power consumption load sequence, the periodicity, seasonality, and influence of special events on power consumption load are considered comprehensively. By establishing a power consumption load prediction model, historical power consumption load data is used as input. After training and optimization of the model, it can output a more accurate predicted power consumption load sequence based on the time characteristics of the preset time zone and relevant influencing factors.

[0042] For the prediction of the power grid frequency sequence, the dynamic relationship between the power grid frequency and the power generation and power consumption load is analyzed. A power grid frequency prediction model is constructed in combination with historical power grid frequency data and the operating characteristics of power generation equipment. The model predicts the trend of the power grid frequency in the preset time zone by considering the inertia, regulation capacity and external interference factors of the power grid.

[0043] When predicting the renewable energy output ratio sequence, the natural characteristics of renewable energy, such as the solar radiation intensity of solar energy and the wind speed of wind power generation, are considered. Historical renewable energy output data and corresponding meteorological data are collected to establish a renewable energy output ratio prediction model based on meteorological factors. By predicting the meteorological conditions in the preset time zone, the model calculates the renewable energy output ratio sequence in the preset time zone.

[0044] Further, after obtaining the above three prediction sequences, the control response speed requirement is determined according to the stability requirement and real-time operating state of the power grid. The control response speed requirement is related to the fluctuation of the power grid. If the predicted power consumption load sequence fluctuates greatly or the predicted power grid frequency sequence deviates from the normal range with a high probability, a faster control response speed is required to ensure the stable operation of the power grid.

[0045] At the same time, the change of the renewable energy output ratio sequence will also affect the control response speed requirement. When the renewable energy output ratio is large and fluctuates frequently, the output power of the distributed resources is adjusted in time to balance the supply and demand relationship of the power grid.

[0046] The historical power grid operating data includes a historical power consumption load sequence, a historical power grid frequency sequence and a historical renewable energy output ratio sequence.

[0047] In the embodiments of the present application, the historical power grid operating data records the operating conditions of the power grid in the past period of time, including the historical power consumption load sequence, the historical power grid frequency sequence and the historical renewable energy output ratio sequence. The historical power consumption load sequence reflects the demand for power by users in different time periods, and by analyzing the periodic and seasonal variation laws, the power consumption peak and valley periods can be grasped.

[0048] The historical power grid frequency sequence reflects the balance between power generation and power consumption in the power grid, and a stable power grid frequency is an index for the safe operation of the power grid. By analyzing the stability and regulation capacity of the power grid, a basis for predicting the trend of future power grid frequency can be provided.

[0049] The historical renewable energy output ratio sequence shows the proportion of renewable energy in the energy structure of the power grid. Due to the intermittent and fluctuating characteristics of renewable energy, the change of the output ratio will affect the operation of the power grid.

[0050] Specifically, the predicted power consumption load sequence, the predicted power grid frequency sequence and the predicted renewable energy output ratio sequence in a preset time zone are predicted according to the historical power grid operation data, and the historical power grid operation data comprises:

[0051] The power consumption load predictor, the power grid frequency predictor and the energy output ratio predictor are respectively constructed based on the long short-term memory network, wherein each predictor comprises L prediction plug-ins, and L is an integer greater than or equal to 10;

[0052] The historical power consumption load sequence, the historical power grid frequency sequence and the historical renewable energy output ratio sequence are respectively subjected to data fluctuation analysis, and the historical power consumption load fluctuation coefficient, the historical power grid frequency fluctuation coefficient and the historical energy output ratio fluctuation coefficient are output;

[0053] The adaptive power consumption load plug-in number, the adaptive power grid frequency plug-in number and the adaptive energy output ratio plug-in number are determined based on the historical power consumption load fluctuation coefficient, the historical power grid frequency fluctuation coefficient and the historical energy output ratio fluctuation coefficient;

[0054] The same number of prediction plug-ins in the power consumption load predictor, the power grid frequency predictor and the energy output ratio predictor are respectively called according to the adaptive power consumption load plug-in number, the adaptive power grid frequency plug-in number and the adaptive energy output ratio plug-in number, and the predicted power consumption load sequence, the predicted power grid frequency sequence and the predicted renewable energy output ratio sequence in a preset time zone are respectively predicted according to the historical power consumption load sequence, the historical power grid frequency sequence and the historical renewable energy output ratio sequence.

[0055] In the embodiment of the application, first, the predictor is constructed based on the long short-term memory network. Since the long short-term memory network can process and predict the long-term dependence relationship in the time series data, it is suitable for the prediction of power grid operation data. Each predictor comprises L prediction plug-ins, and L is an integer greater than or equal to 10. The prediction plug-ins can process data in parallel, improving the efficiency and accuracy of prediction.

[0056] Exemplarily, the power consumption load predictor is constructed based on the long short-term memory network, and the specific steps are as follows:

[0057] First, data acquisition, the historical power consumption load sequence is acquired based on the historical power grid operation data.

[0058] Secondly, the model is built, taking the historical electricity consumption sequence as input and the predicted electricity load as output. The input layer has a node number equal to the dimension of the input features, such as 2 features in the historical electricity consumption sequence, so the input layer contains 2 nodes; 1-3 hidden layers are set, and the number of nodes in each layer is adjusted through experiments, such as 64, 32, etc., and the activation function is selected as ReLU; the number of nodes in the output layer is equal to the number of predicted electricity load, such as 1 node for prediction time, and the output layer generally does not use the activation function, directly outputting continuous values.

[0059] Then, the model is trained, and in each training iteration, the Adam optimizer and the mean square error (MSE) loss function are used to build the training framework, the batch size is set to 32, the total training rounds are set to 50, and the early stopping mechanism is introduced, patience = 5, when the validation set loss does not decrease for 5 consecutive rounds, the training process is automatically terminated, and the trained electricity load predictor is obtained.

[0060] Similarly, the electricity load predictor and the energy output ratio predictor are built according to the above method.

[0061] Secondly, the historical electricity load sequence, the historical grid frequency sequence and the historical renewable energy output ratio sequence are analyzed for data volatility, such as by calculating the ratio of the standard deviation to the mean of the data sequence. The historical electricity load fluctuation coefficient, the historical grid frequency fluctuation coefficient and the historical energy output ratio fluctuation coefficient can reflect the degree of data fluctuation, and then the number of adaptive plug-ins is determined according to the fluctuation coefficient. Data with larger fluctuations may require more plug-ins for accurate prediction, while data with smaller fluctuations can use relatively fewer plug-ins.

[0062] Thirdly, after determining the number of adaptive electricity load plug-ins, the number of adaptive grid frequency plug-ins and the number of adaptive energy output ratio plug-ins, the same number of prediction plug-ins in the electricity load predictor, the grid frequency predictor and the energy output ratio predictor are called respectively. The prediction plug-in will use the algorithm and model of the long short-term memory network to predict the relevant sequence in the preset time zone according to the input historical electricity load sequence, historical grid frequency sequence and historical renewable energy output ratio sequence.

[0063] Further, in the prediction process, the prediction plug-in improves the accuracy of the prediction by continuously learning and adjusting the model parameters. At the same time, various factors that may affect the prediction results are considered, such as weather changes, holidays and other special circumstances. Through the above method, the predicted electricity load sequence, the predicted grid frequency sequence and the predicted renewable energy output ratio sequence are obtained, which provide reliable data support for the subsequent operation and control scheme.

[0064] After obtaining the predicted sequence and determining the demand for speed of response of regulation, the next step is to optimize the operation and regulation scheme. In the simulation space of the power plant, according to the predicted sequence and the demand for speed of response of regulation, in combination with the instruction tracking error index, the regulation performance index and the power grid influence index and other aspects, different operation and regulation schemes are evaluated and compared. Through continuous optimization and adjustment of the scheme, the optimal operation and regulation scheme is found, the efficient and stable operation of the virtual power plant is realized, and the stability of the power grid operation is further improved, and the risk of power supply is reduced.

[0065] The number of adaptive electricity load plug-ins, the number of adaptive power grid frequency plug-ins and the number of adaptive energy output ratio plug-ins are determined based on the historical electricity load fluctuation coefficient, the historical power grid frequency fluctuation coefficient and the historical energy output ratio fluctuation coefficient, comprising:

[0066] The number of adaptive electricity load plug-ins is obtained by multiplying the ratio of the historical electricity load fluctuation coefficient to the maximum historical electricity load fluctuation coefficient in the historical time range by L and taking the integer.

[0067] The number of adaptive power grid frequency plug-ins is obtained by multiplying the ratio of the historical power grid frequency fluctuation coefficient to the maximum historical power grid frequency fluctuation coefficient in the historical time range by L and taking the integer.

[0068] The number of adaptive energy output ratio plug-ins is obtained by multiplying the ratio of the historical energy output ratio fluctuation coefficient to the maximum historical energy output ratio fluctuation coefficient in the historical time range by L and taking the integer.

[0069] In the embodiment of the application, first, the ratio of the historical electricity load fluctuation coefficient to the maximum historical electricity load fluctuation coefficient in the historical time range is calculated, which reflects the degree of the current historical electricity load fluctuation relative to the historical maximum fluctuation. The ratio is multiplied by the total number L of prediction plug-ins in the predictor and is taken as an integer to obtain the number of adaptive electricity load plug-ins.

[0070] For example, if the ratio is 0.6 and L is 20, then the number of adaptive electricity load plug-ins is 0.6x20=12. Since the greater the fluctuation degree, more prediction plug-ins are needed to capture the change rule of the data to improve the accuracy of prediction.

[0071] Similarly, for the historical power grid frequency fluctuation coefficient, the ratio of the historical power grid frequency fluctuation coefficient to the maximum historical power grid frequency fluctuation coefficient in the historical time range is calculated. The ratio reflects the relative size of the current historical power grid frequency fluctuation. The ratio is multiplied by L and taken as an integer to obtain the number of adaptive power grid frequency plug-ins.

[0072] Exemplarily, if the ratio is 0.7 and L is 20, the number of adaptive grid frequency plug-ins is 0.7*20=14. Since the fluctuation of the grid frequency has a greater impact on the stable operation of the grid, more plug-ins are needed to accurately predict the changes when the fluctuation is large, so as to better cope with possible frequency abnormal situations and ensure the stable operation of the grid.

[0073] The historical energy output ratio fluctuation coefficient is similar to the above, and the ratio of the historical energy output ratio fluctuation coefficient to the maximum historical energy output ratio fluctuation coefficient in the historical time range is first calculated. This ratio reflects the relative situation of the fluctuation of the renewable energy output ratio. Multiply this ratio by L and take the integer to obtain the number of adaptive energy output ratio plug-ins.

[0074] Exemplarily, if the ratio of the historical energy output ratio fluctuation coefficient to the maximum historical energy output ratio fluctuation coefficient is 0.8 and L is 20, then the number of adaptive energy output ratio plug-ins is 0.8*20=16. Since renewable energy has intermittency and volatility, the fluctuation of its output ratio will affect the balance of supply and demand of the grid. Therefore, according to the fluctuation degree, the number of prediction plug-ins is reasonably allocated, which helps to more accurately predict the renewable energy output ratio sequence, and then realize the effective allocation of distributed resources.

[0075] By determining the number of adaptive plug-ins in the above manner, the prediction resources can be reasonably allocated according to the fluctuation characteristics of different data, so that the predictor can achieve good prediction results when processing different types of historical grid operation data, and provide a more reliable basis for subsequent virtual power plant operation and control scheme formulation. After determining the number of adaptive plug-ins, the prediction plug-ins in the predictor are called according to the corresponding number to accurately predict the electricity load, grid frequency and renewable energy output ratio in the preset time zone, further laying a foundation for the efficient and stable operation of the virtual power plant.

[0076] Further, the control response speed demand is determined, including:

[0077] According to the predicted electricity load sequence, the load change gradient and the maximum climbing rate are calculated and identified to obtain the predicted load change gradient and the predicted maximum climbing rate;

[0078] The frequency fluctuation of the predicted grid frequency sequence is calculated, and the predicted grid frequency fluctuation coefficient is output;

[0079] The Fourier transform analysis is performed on the predicted grid frequency sequence, the frequency band with the highest power spectral density is identified, and the dominant fluctuation frequency range is determined;

[0080] According to the predicted renewable energy output ratio sequence, the system structure vulnerability is evaluated, and the system vulnerability coefficient is determined;

[0081] The control response speed requirement is determined based on the predicted load change gradient, the predicted maximum ramp rate, the predicted power grid frequency fluctuation coefficient, the dominant fluctuation frequency range, and the system vulnerability coefficient.

[0082] In the embodiments of the present application, first, the load change gradient is calculated according to the predicted power load sequence, and the maximum ramp rate is identified. The load change gradient reflects the rate of change of the power load with time, and can reflect the speed of the power load change. By calculating and analyzing the load difference between adjacent time points in the predicted power load sequence, the load change gradient is obtained. The maximum ramp rate refers to the maximum rising or falling rate of the power load in a short time. By point-by-point analysis of the predicted power load sequence, the time period with the fastest load change is found, thereby identifying the maximum ramp rate, obtaining the predicted load change gradient and the predicted maximum ramp rate.

[0083] Secondly, the frequency fluctuation of the predicted power grid frequency sequence is calculated. Power grid frequency fluctuation can cause damage to power equipment and affect the stable operation of the power system. By calculating the ratio of the standard deviation to the mean value of the data in the predicted power grid frequency sequence, the predicted power grid frequency fluctuation coefficient is output, which can intuitively reflect the fluctuation degree of the power grid frequency.

[0084] For example, if the predicted power grid frequency sequence is [50.1, 49.9, 50.2, 49.8], the mean value is about 50, and the standard deviation is about 0.17, then the predicted power grid frequency fluctuation coefficient is about 0.17 ÷ 50 = 0.0034.

[0085] Then, the predicted power grid frequency sequence is analyzed by Fourier transform. Fourier transform can convert the signal in the time domain to the signal in the frequency domain. By analyzing the signal characteristics in the frequency domain, the frequency band with the highest power spectral density is identified. The frequency band with high power spectral density represents a signal with strong energy in that frequency band, which has a greater impact on the power grid frequency fluctuation.

[0086] Thirdly, the system structure vulnerability degree of the predicted renewable energy output ratio sequence is evaluated. Due to the intermittent and fluctuating characteristics of renewable energy, the change of the output ratio of renewable energy will affect the system structure of the power grid. By establishing a system structure vulnerability degree evaluation model, considering factors such as the distribution of renewable energy, the access point, and the complementary relationship with traditional energy, the system vulnerability coefficient is determined. The higher the system vulnerability coefficient, the more vulnerable the power grid is to the change of renewable energy output, and the poorer the stability.

[0087] Finally, the demand for the speed of the regulation response is determined based on the predicted load change gradient, the predicted maximum ramp rate, the predicted power grid frequency fluctuation coefficient, the dominant fluctuation frequency range, and the system vulnerability coefficient. When the predicted load change gradient is large, the predicted maximum ramp rate is high, the predicted power grid frequency fluctuation coefficient is large, the dominant fluctuation frequency range is unstable, or the system vulnerability coefficient is high, it indicates that the operation state of the power grid is unstable, and a faster speed of the regulation response is needed to ensure that the power grid can respond to various changes in time and maintain stable operation.

[0088] Specifically, the predicted load change gradient, the predicted maximum ramp rate, the predicted power grid frequency fluctuation coefficient, the dominant fluctuation frequency range, and the system vulnerability coefficient are normalized, and the demand for the speed of the regulation response is determined through intelligent evaluation by the power grid expert system. The demand for the speed of the regulation response is positively correlated with the predicted load change gradient, the predicted maximum ramp rate, the predicted power grid frequency fluctuation coefficient, the dominant fluctuation frequency range, and the system vulnerability coefficient.

[0089] In the embodiments of the present application, the normalization processing can eliminate the influence of the dimensions and orders of magnitude between different indicators, so that each indicator is comparable. After the predicted load change gradient, the predicted maximum ramp rate, the predicted power grid frequency fluctuation coefficient, the dominant fluctuation frequency range, and the system vulnerability coefficient are normalized, the above indicators are mapped to a unified interval, which facilitates subsequent comprehensive evaluation.

[0090] The power grid expert system is an existing artificial intelligence model that contains a large amount of knowledge and experience of power grid operation and regulation. The normalized indicators are intelligently evaluated by the power grid expert system, and the demand for the speed of the regulation response is determined in combination with the actual operation situation and historical data of the power grid. Since the demand for the speed of the regulation response is positively correlated with the above indicators, when the values of the indicators are larger, the expert system will determine that a faster speed of the regulation response is needed.

[0091] For example, if the predicted load change gradient is large, it means that the electricity load changes rapidly in a short time. In order to ensure the balance between supply and demand of the power grid, it is necessary to quickly adjust the output power of the distributed resources, which requires a faster speed of the regulation response. Similarly, when the predicted maximum ramp rate is high, the predicted power grid frequency fluctuation coefficient is large, the dominant fluctuation frequency range is unstable, or the system vulnerability coefficient is high, it indicates that the power grid faces a greater operation risk, and timely and fast regulation is needed to maintain stability.

[0092] Through the evaluation method based on the normalized indicators and the power grid expert system, the demand for the speed of the regulation response is determined, which provides key parameters for the subsequent formulation and implementation of the virtual power plant operation and regulation scheme, so as to realize effective regulation of the virtual power plant on the operation of the power grid and ensure the safe, stable, and efficient operation of the power grid.

[0093] S30: Determine an adaptive resource combination strategy according to the demand for the speed of response to regulation, and perform optimization of an operation and regulation scheme in the power plant simulation space based on the adaptive resource combination strategy to maximize the accuracy of regulation, and determine an optimal regulation scheme for operation and regulation.

[0094] In the embodiments of the present application, the adaptive resource combination strategy is determined after the demand for the speed of response to regulation is determined, and can be obtained based on historical data by constructing a response speed-resource comparison table. The adaptive resource combination strategy needs to comprehensively consider the characteristics and capabilities of various distributed resources, including but not limited to distributed power sources, energy storage devices, and controllable loads. Different distributed resources differ in response speed, regulation capacity, and cost, and therefore need to be reasonably matched and combined according to the demand for the speed of response to regulation.

[0095] In the power plant simulation space, optimization of an operation and regulation scheme is performed by frequency modulation power ratio adjustment based on the adaptive resource combination strategy to maximize the accuracy of regulation. An accurate simulation model is established, considering the power grid topology, device parameters, operation constraints, and the like. By simulating and analyzing different operation and regulation schemes, the accuracy and effect of regulation are evaluated.

[0096] In the optimization process, an optimization algorithm such as a genetic algorithm or a particle swarm algorithm can be used to search for an optimal operation and regulation scheme. The optimization algorithm can quickly find a better solution in a complex solution space, improving the efficiency and accuracy of optimization. At the same time, the optimization results can be verified and adjusted in combination with the actual operation of the power grid and historical data to ensure the feasibility and effectiveness of the optimal regulation scheme.

[0097] After the optimal regulation scheme is determined, operation and regulation can be performed. In the operation and regulation process, the operation state of the power grid needs to be monitored in real time, and the regulation scheme needs to be dynamically adjusted according to the actual situation. At the same time, coordination and communication need to be carried out with the owners and operators of distributed resources to ensure the smooth execution of regulation instructions.

[0098] In the power plant simulation space, optimization of an operation and regulation scheme is performed based on the adaptive resource combination strategy to maximize the accuracy of regulation, including:

[0099] Determine a plurality of power ratio adjustment thresholds for a plurality of adaptive resources based on the adaptive resource combination strategy, and construct a frequency modulation power ratio adjustment space;

[0100] Randomly select power ratio values in the frequency modulation power ratio adjustment space to obtain a first initial regulation scheme;

[0101] In the power plant simulation space, according to the first initial regulation scheme, the predicted power load sequence, the predicted power grid frequency sequence and the predicted renewable energy output ratio sequence, regulation simulation is performed, and a first simulated regulation accuracy is output.

[0102] The regulation scheme iteration selection and iteration evaluation continue to be performed in the frequency modulation power ratio adjustment space until a preset convergence number is reached, and the initial regulation scheme corresponding to the maximum simulated regulation accuracy is set as an optimal regulation scheme.

[0103] In the embodiments of the present application, firstly, a plurality of power ratio adjustment thresholds of a plurality of adaptive resources are determined based on an adaptive resource combination strategy. Different distributed resources, such as distributed power sources, energy storage devices and controllable loads, have different roles in the operation and regulation of the virtual power plant, so the power ratio adjustment thresholds need to be determined according to their characteristics and regulation response speed requirements.

[0104] For example, for energy storage devices with fast response speed, a larger power ratio adjustment range is set; and for distributed power sources with slow response speed but stable power generation, a relatively smaller adjustment range is set.

[0105] Secondly, power ratio values are randomly selected in the frequency modulation power ratio adjustment space to obtain a first initial regulation scheme. The random selection method can ensure the diversity of the initial scheme and avoid falling into a local optimal solution. Each power ratio value represents the power distribution ratio of different adaptive resources in the regulation process, and the combination of the ratio values forms a specific regulation scheme.

[0106] Then, in the power plant simulation space, regulation simulation is performed according to the first initial regulation scheme, the predicted power load sequence, the predicted power grid frequency sequence and the predicted renewable energy output ratio sequence. The simulation space considers various factors such as power grid topology, device parameters and operation constraints to establish an accurate power grid model. By applying the first initial regulation scheme to the simulation model and combining the predicted sequence data, the regulation process of the power grid under the scheme is simulated, and a first simulated regulation accuracy is output. The accuracy index reflects the regulation effect of the regulation scheme on the power grid operation under the current predicted data.

[0107] After that, regulation scheme iteration selection and iteration evaluation continue to be performed in the frequency modulation power ratio adjustment space. In each iteration, new power ratio values are selected from the frequency modulation power ratio adjustment space to form a new regulation scheme, and simulation and evaluation are performed in the power plant simulation space to obtain a new simulated regulation accuracy. The above process is repeatedly performed until a preset convergence number is reached. Setting the convergence number ensures that the optimization process ends within a reasonable time, and also ensures that the optimal scheme found has high accuracy to a certain extent.

[0108] Finally, the initial control scheme corresponding to the maximum simulation control precision is set as the optimal control scheme. The optimal control scheme considers various prediction data and adaptive resource combination strategies, and is obtained through multiple iterations, thereby maximizing the control precision in the power plant simulation space, and providing a reliable basis for the actual operation and control of the virtual power plant.

[0109] In actual operation and control, the optimal control scheme is followed, and the power grid operation state is monitored in real time, and dynamic adjustment is made according to the actual situation, so as to ensure the efficient and stable operation of the virtual power plant and the safe and reliable power supply of the power grid.

[0110] Further, the first simulation control precision is output, including:

[0111] According to the first initial control scheme, the predicted power load sequence, the predicted power grid frequency sequence and the predicted renewable energy output ratio sequence, a control simulation is performed, and a first control simulation result is output, wherein the first control simulation result includes a first instruction tracking error index, a first regulation performance index and a first power grid influence index, the first instruction tracking error index includes a root mean square error and an average absolute error, the first regulation performance index includes an average response delay and a regulation time, and the first power grid influence index includes a power grid frequency standard deviation improvement rate.

[0112] According to the first instruction tracking error index, the first regulation performance index and the first power grid influence index, a control precision evaluation is performed, and a first simulation control precision is output.

[0113] In the embodiments of the present application, first, according to the first initial control scheme, the predicted power load sequence, the predicted power grid frequency sequence and the predicted renewable energy output ratio sequence, a control simulation is performed. In the simulation process, the first initial control scheme is applied to the power grid simulation model, and through the operation simulation of the model, relevant data are recorded to obtain the first control simulation result.

[0114] For the first instruction tracking error index, the root mean square error reflects the square root of the average value of the square of the error between the control instruction and the actual execution, and can measure the overall deviation degree of the control instruction tracking. The average absolute error is the average value of the absolute value of the error, which reflects the average deviation size of each control instruction execution. Through analysis and calculation of the control instruction and the actual response data in the simulation process, the root mean square error and the average absolute error are obtained.

[0115] Specifically, the average response delay in the first adjustment performance index refers to the average time interval from issuing the control instruction to the start of the response of the distributed resource, reflecting the timeliness of the response of the distributed resource to the control instruction. The adjustment time refers to the time required from the start of the control to the system reaching a stable state, reflecting the rapidity and stability of the control process. In the simulation process, the response time of each distributed resource and the time for the system to reach a stable state are recorded, and then the average response delay and the adjustment time are calculated.

[0116] The improvement rate of the standard deviation of the grid frequency in the first grid influence index is measured by comparing the standard deviation of the grid frequency before and after the control. The standard deviation reflects the fluctuation degree of the grid frequency, and the improvement rate reflects the improvement effect of the control scheme on the stability of the grid frequency. The standard deviation of the grid frequency is calculated before and after the simulation control, and then the improvement rate is calculated.

[0117] For example, assuming that the standard deviation of the grid frequency before the control is 0.2 and the standard deviation of the grid frequency after the control is 0.15, then the improvement rate of the standard deviation of the grid frequency is (0.2-0.15) ÷ 0.2 = 0.25, i.e. 25%, indicating that the control scheme reduces the fluctuation degree of the grid frequency by 25%, and has a certain improvement effect on the stability of the grid frequency.

[0118] Secondly, the control precision is evaluated according to the first instruction tracking error index, the first adjustment performance index and the first grid influence index, and the first simulation control precision is output. For example, different weights are assigned to each index, and the control precision is evaluated by weighted average.

[0119] For example, the root mean square error is assigned a weight of 0.3, the average absolute error is assigned a weight of 0.2, the average response delay is assigned a weight of 0.2, the adjustment time is assigned a weight of 0.1, and the improvement rate of the standard deviation of the grid frequency is assigned a weight of 0.2. The values of each index are multiplied by the corresponding weight and added together, and the result is the first simulation control precision.

[0120] Through the above simulation and evaluation process, the effect of each control scheme is evaluated, so that the optimal control scheme is selected from a large number of schemes, providing a solid foundation for the operation and control of the virtual power plant, and ensuring that the grid can operate stably and efficiently under various complex conditions.

[0121] In summary, compared with the prior art, the present application first models according to the controllable characteristics of the equipment to accurately analyze the balance relationship between the output boundary of the combined heat and power unit, the charging and discharging power and state of charge constraint of the energy storage, and the system power; then generates the equipment output plan and grid interaction scheme based on the model combined with the source and load operation data; finally realizes the model parameter configuration, control instruction issuing and running state monitoring through the device, thereby realizing the accurate control of the multiple types of distributed resources in the virtual power plant.

[0122] To sum up, the embodiments of the present application have at least the following technical effects:

[0123] The embodiments of the present application provide a virtual power plant operation regulation method adaptive to multiple distributed resources. First, a power plant simulation space is constructed on a three-dimensional simulation platform to comprehensively and accurately simulate the actual operation environment of the virtual power plant, laying a solid foundation for subsequent regulation work. Second, historical power grid operation data is obtained through monitoring and metering, and then a power consumption load predictor, a power grid frequency predictor, and an energy output ratio predictor are constructed, which can more accurately predict the predicted power consumption load sequence, the predicted power grid frequency sequence, and the predicted renewable energy output ratio sequence in a preset time zone. Finally, a resource combination strategy adaptive to the regulation response speed requirement is determined, and an operation regulation scheme optimization is performed in the power plant simulation space, comprehensively considering multiple aspects such as instruction tracking error index, regulation performance index, and power grid influence index, to ensure the comprehensiveness and accuracy of regulation precision evaluation. Through the above technical solutions, the method effectively solves the problem that the existing virtual power plant operation regulation method cannot accurately depict the operation constraints of multiple types of distributed resources in the power plant, and the regulation strategy is disconnected from the actual device characteristics, improves the virtual power plant operation regulation efficiency, enhances the stability of the power grid operation, reduces the risk of power supply, and provides protection for the efficient operation of the virtual power plant.

[0124] In one embodiment, the virtual power plant operation regulation method adaptive to multiple distributed resources comprises the following steps: Figure 2 According to the same inventive concept of the virtual power plant operation regulation method adaptive to multiple distributed resources provided in the first embodiment, the embodiments of the present application further provide a virtual power plant operation regulation device adaptive to multiple distributed resources, which comprises:

[0125] The space construction module 11 is configured to construct a power plant simulation space comprising a distributed resource layer, a monitoring and metering layer, and an operation regulation layer in a three-dimensional simulation platform.

[0126] The sequence prediction module 12 is configured to obtain historical power grid operation data in a historical time zone through the monitoring and metering layer, predict a predicted power consumption load sequence, a predicted power grid frequency sequence, and a predicted renewable energy output ratio sequence in a preset time zone according to the historical power grid operation data, and determine a regulation response speed requirement.

[0127] The scheme execution module 13 is configured to determine a resource combination strategy adaptive to the regulation response speed requirement, perform operation regulation scheme optimization in the power plant simulation space based on the resource combination strategy adaptive to the regulation response speed requirement, and determine an optimal regulation scheme for operation regulation.

[0128] In one embodiment, the space construction module 11 is specifically configured to:

[0129] constructing a distributed resource layer in the three-dimensional simulation platform, wherein the distributed resource layer comprises a combined heat and power unit, an energy storage system, a wind power generation unit and a photovoltaic power generation unit, and each type of resource comprises a corresponding operation mathematical model, the operation mathematical model at least containing a ramp rate, a response delay, an energy state and an operation boundary constraint;

[0130] in the three-dimensional simulation platform, simulating an intelligent monitoring and metering device deployed on each distributed resource to collect real-time operation state data and construct a monitoring and metering layer;

[0131] constructing an operation regulation layer in the three-dimensional simulation platform, wherein the operation regulation layer is configured to receive data from the monitoring and metering layer and perform subsequent operation regulation;

[0132] building a power plant simulation space based on the distributed resource layer, the monitoring and metering layer and the operation regulation layer.

[0133] The historical power grid operation data comprises a historical power consumption load sequence, a historical power grid frequency sequence and a historical renewable energy output proportion sequence.

[0134] Further, in an application embodiment, the predicted power consumption load sequence, the predicted power grid frequency sequence and the predicted renewable energy output proportion sequence in a preset time zone are predicted according to the historical power grid operation data, comprising:

[0135] building a power consumption load predictor, a power grid frequency predictor and an energy output proportion predictor based on a long short-term memory network, wherein each predictor comprises L prediction plug-ins, and L is an integer greater than or equal to 10;

[0136] performing data volatility analysis on the historical power consumption load sequence, the historical power grid frequency sequence and the historical renewable energy output proportion sequence, and outputting a historical power consumption load volatility coefficient, a historical power grid frequency volatility coefficient and a historical energy output proportion volatility coefficient;

[0137] determining an adaptive power consumption load plug-in number, an adaptive power grid frequency plug-in number and an adaptive energy output proportion plug-in number based on the historical power consumption load volatility coefficient, the historical power grid frequency volatility coefficient and the historical energy output proportion volatility coefficient;

[0138] According to the adaptive power consumption load plug-in number, the adaptive power grid frequency plug-in number and the adaptive energy output proportion plug-in number, the same number of prediction plug-ins in the power consumption load predictor, the power grid frequency predictor and the energy output proportion predictor are called respectively, and the predicted power consumption load sequence, the predicted power grid frequency sequence and the predicted renewable energy output proportion sequence in a preset time zone are predicted respectively according to the historical power consumption load sequence, the historical power grid frequency sequence and the historical renewable energy output proportion sequence.

[0139] Further, in one application embodiment, based on the historical electricity load fluctuation coefficient, the historical power grid frequency fluctuation coefficient and the historical energy output proportion fluctuation coefficient, the adaptive electricity load plug-in quantity, the adaptive power grid frequency plug-in quantity and the adaptive energy output proportion plug-in quantity are determined, comprising:

[0140] The ratio of the historical electricity load fluctuation coefficient to the maximum historical electricity load fluctuation coefficient in the historical time range is multiplied by L to obtain the adaptive electricity load plug-in quantity;

[0141] The ratio of the historical power grid frequency fluctuation coefficient to the maximum historical power grid frequency fluctuation coefficient in the historical time range is multiplied by L to obtain the adaptive power grid frequency plug-in quantity;

[0142] The ratio of the historical energy output proportion fluctuation coefficient to the maximum historical energy output proportion fluctuation coefficient in the historical time range is multiplied by L to obtain the adaptive energy output proportion plug-in quantity.

[0143] Determine the control response speed demand, comprising:

[0144] According to the predicted electricity load sequence, the load change gradient is calculated and the maximum climbing rate is identified, and the predicted load change gradient and the predicted maximum climbing rate are obtained;

[0145] The predicted power grid frequency sequence is calculated for frequency fluctuation, and the predicted power grid frequency fluctuation coefficient is output;

[0146] The predicted power grid frequency sequence is analyzed by Fourier transform, the frequency band with the highest power spectral density is identified, and the dominant fluctuation frequency range is determined;

[0147] According to the predicted renewable energy output proportion sequence, the system structure vulnerability is evaluated, and the system vulnerability coefficient is determined;

[0148] Based on the predicted load change gradient, the predicted maximum climbing rate, the predicted power grid frequency fluctuation coefficient, the dominant fluctuation frequency range and the system vulnerability coefficient, the control response speed demand is determined.

[0149] Further, the predicted load change gradient, the predicted maximum climbing rate, the predicted power grid frequency fluctuation coefficient, the dominant fluctuation frequency range and the system vulnerability coefficient are normalized, and the control response speed demand is determined by intelligent evaluation of the power grid expert system, wherein the control response speed demand and the predicted load change gradient, the predicted maximum climbing rate, the predicted power grid frequency fluctuation coefficient, the dominant fluctuation frequency range and the system vulnerability coefficient are positively correlated.

[0150] Further, in the power plant simulation space, the operation regulation scheme optimization is performed based on the adaptive resource combination strategy to maximize the regulation accuracy, including:

[0151] A plurality of power ratio adjustment thresholds of the adaptive resources are determined based on the adaptive resource combination strategy, and a frequency regulation power ratio adjustment space is constructed.

[0152] A power ratio value is randomly selected in the frequency regulation power ratio adjustment space to obtain a first initial regulation scheme.

[0153] In the power plant simulation space, the regulation simulation is performed according to the first initial regulation scheme, the predicted power load sequence, the predicted power grid frequency sequence, and the predicted renewable energy output ratio sequence, and a first simulation regulation accuracy is output.

[0154] The regulation scheme iteration selection and iteration evaluation continue in the frequency regulation power ratio adjustment space until a preset convergence number is reached, and the initial regulation scheme corresponding to the maximum simulation regulation accuracy is set as the optimal regulation scheme.

[0155] Further, in one embodiment, the first simulation regulation accuracy is output, including:

[0156] The regulation simulation is performed according to the first initial regulation scheme, the predicted power load sequence, the predicted power grid frequency sequence, and the predicted renewable energy output ratio sequence, and a first regulation simulation result is output, wherein the first regulation simulation result includes a first instruction tracking error index, a first regulation performance index, and a first power grid influence index, the first instruction tracking error index includes a root mean square error and an average absolute error, the first regulation performance index includes an average response delay and a regulation time, and the first power grid influence index includes a power grid frequency standard deviation improvement rate.

[0157] The regulation accuracy is evaluated according to the first instruction tracking error index, the first regulation performance index, and the first power grid influence index, and the first simulation regulation accuracy is output.

[0158] It should be noted that the above sequence of embodiments of the present application is only for description, and does not represent the advantages and disadvantages of the embodiments. The above describes a specific embodiment of the present application. In addition, the processes depicted in the drawings do not necessarily require the specific order or continuous order shown to achieve the desired results. In some embodiments, multi-task processing and parallel processing are possible or may be advantageous.

[0159] The above only describes the preferred embodiments of the present application and does not limit the present application. Any modification, equivalent replacement, improvement, etc. made within the spirit and principles of the present application shall be included in the protection scope of the present application.

[0160] The specification and drawings are, of course, to be regarded in an illustrative rather than a restrictive sense. It is to be understood that any such modifications, variations, combinations or equivalents that fall within the scope of the application are intended to be embraced herein.

Claims

1. A virtual power plant operation regulation method for adapting to multiple distributed resources, characterized in that, The method comprises: constructing a power plant simulation space comprising a distributed resource layer, a monitoring and metering layer, and a running and regulating layer in a three-dimensional simulation platform; monitoring and obtaining historical power grid operation data in a historical time zone through the monitoring and metering layer, predicting a predicted power consumption load sequence, a predicted power grid frequency sequence, and a predicted renewable energy output ratio sequence in a preset time zone according to the historical power grid operation data, and determining a regulating response speed requirement; determining an adaptive resource combination strategy according to the regulating response speed requirement, and performing running and regulating scheme optimization in the power plant simulation space based on the adaptive resource combination strategy to maximize the regulating accuracy, and determining an optimal regulating scheme for running and regulating; wherein constructing a power plant simulation space comprising a distributed resource layer, a monitoring and metering layer, and a running and regulating layer in a three-dimensional simulation platform comprises: constructing a distributed resource layer in a three-dimensional simulation platform, wherein the distributed resource layer comprises a combined heat and power unit, an energy storage system, a wind power generation unit, and a photovoltaic power generation unit, and each type of resource comprises a corresponding running mathematical model, and the running mathematical model at least comprises a ramp rate, a response delay, an energy state, and an operating boundary constraint; in a three-dimensional simulation platform, simulating and deploying intelligent monitoring and metering devices on each distributed resource to collect running state data in real time, and constructing a monitoring and metering layer; constructing a running and regulating layer in a three-dimensional simulation platform, wherein the running and regulating layer is used to receive data of the monitoring and metering layer and perform subsequent running and regulating; building a power plant simulation space based on the distributed resource layer, the monitoring and metering layer, and the running and regulating layer; wherein, in the power plant simulation space, performing running and regulating scheme optimization based on the adaptive resource combination strategy to maximize the regulating accuracy comprises: determining a plurality of power ratio adjustment thresholds of a plurality of adaptive resources based on the adaptive resource combination strategy to construct a frequency modulation power ratio adjustment space; randomly selecting power ratio values in the frequency modulation power ratio adjustment space to combine to obtain a first initial regulating scheme; in the power plant simulation space, performing regulating simulation according to the first initial regulating scheme, the predicted power consumption load sequence, the predicted power grid frequency sequence, and the predicted renewable energy output ratio sequence, and outputting a first simulated regulating accuracy; continuing to perform regulating scheme iteration selection and iteration evaluation in the frequency modulation power ratio adjustment space until a preset convergence number is reached, and setting an initial regulating scheme corresponding to the maximum simulated regulating accuracy as an optimal regulating scheme. 2.The virtual power plant operation regulation method for adapting multiple distributed resources according to claim 1, wherein, monitoring and obtaining historical power grid operation data in a historical time zone through the monitoring and metering layer, wherein the historical power grid operation data comprises a historical power consumption load sequence, a historical power grid frequency sequence, and a historical renewable energy output ratio sequence. 3.The virtual power plant operation regulation method for adapting multiple distributed resources according to claim 2, wherein, predicting a predicted power consumption load sequence, a predicted power grid frequency sequence, and a predicted renewable energy output ratio sequence in a preset time zone according to the historical power grid operation data, comprising: building a power consumption load predictor, a power grid frequency predictor, and an energy output ratio predictor based on a long short-term memory network, wherein each predictor comprises L prediction plugins, and L is an integer greater than or equal to 10; respectively, the historical electricity load fluctuation coefficient, the historical power grid frequency fluctuation coefficient and the historical energy output ratio fluctuation coefficient are determined; The number of adaptive electricity load plug-ins, the number of adaptive power grid frequency plug-ins and the number of adaptive energy output ratio plug-ins are determined based on the historical electricity load fluctuation coefficient, the historical power grid frequency fluctuation coefficient and the historical energy output ratio fluctuation coefficient. The same number of prediction plug-ins in the electricity load predictor, the power grid frequency predictor and the energy output ratio predictor are called respectively according to the number of adaptive electricity load plug-ins, the number of adaptive power grid frequency plug-ins and the number of adaptive energy output ratio plug-ins, and the predicted electricity load sequence, the predicted power grid frequency sequence and the predicted renewable energy output ratio sequence in the preset time zone are respectively predicted based on the historical electricity load sequence, the historical power grid frequency sequence and the historical renewable energy output ratio sequence. 4.The virtual power plant operation regulation method for adapting multiple distributed resources according to claim 3, wherein, The number of adaptive electricity load plug-ins, the number of adaptive power grid frequency plug-ins and the number of adaptive energy output ratio plug-ins are determined based on the historical electricity load fluctuation coefficient, the historical power grid frequency fluctuation coefficient and the historical energy output ratio fluctuation coefficient. The number of adaptive electricity load plug-ins is obtained by multiplying the ratio of the historical electricity load fluctuation coefficient to the maximum historical electricity load fluctuation coefficient in the historical time range by L and taking the integer part. The number of adaptive power grid frequency plug-ins is obtained by multiplying the ratio of the historical power grid frequency fluctuation coefficient to the maximum historical power grid frequency fluctuation coefficient in the historical time range by L and taking the integer part. The number of adaptive energy output ratio plug-ins is obtained by multiplying the ratio of the historical energy output ratio fluctuation coefficient to the maximum historical energy output ratio fluctuation coefficient in the historical time range by L and taking the integer part. 5.The virtual power plant operation regulation method for adapting multiple distributed resources according to claim 1, wherein, The regulation response speed demand is determined, including: The load change gradient and the maximum climbing rate are calculated according to the predicted electricity load sequence, and the predicted load change gradient and the predicted maximum climbing rate are obtained. The frequency fluctuation of the predicted power grid frequency sequence is calculated, and the predicted power grid frequency fluctuation coefficient is output. Fourier transform analysis is performed on the predicted power grid frequency sequence, the frequency band with the highest power spectral density is identified, and the dominant fluctuation frequency range is determined. The system structure vulnerability assessment is performed according to the predicted renewable energy output ratio sequence, and the system vulnerability coefficient is determined. The regulation response speed demand is determined based on the predicted load change gradient, the predicted maximum climbing rate, the predicted power grid frequency fluctuation coefficient, the dominant fluctuation frequency range and the system vulnerability coefficient. 6.The virtual power plant operation regulation method for adapting multiple distributed resources according to claim 5, wherein, The predicted load change gradient, the predicted maximum climbing rate, the predicted power grid frequency fluctuation coefficient, the dominant fluctuation frequency range and the system vulnerability coefficient are normalized and intelligently evaluated by the power grid expert system to determine the regulation response speed demand, wherein the regulation response speed demand is positively correlated with the predicted load change gradient, the predicted maximum climbing rate, the predicted power grid frequency fluctuation coefficient, the dominant fluctuation frequency range and the system vulnerability coefficient. 7.The virtual power plant operation regulation method for adapting multiple distributed resources according to claim 1, wherein, The first simulation regulation accuracy is output, including: According to the first initial regulation scheme, the predicted power load sequence, the predicted power grid frequency sequence, and the predicted renewable energy output ratio sequence, a regulation simulation is performed, and a first regulation simulation result is output, wherein the first regulation simulation result includes a first instruction tracking error index, a first regulation performance index, and a first power grid impact index, the first instruction tracking error index includes a root mean square error and a mean absolute error, the first regulation performance index includes an average response delay and a regulation time, and the first power grid impact index includes a power grid frequency standard deviation improvement rate; According to the first instruction tracking error index, the first regulation performance index, and the first power grid impact index, a regulation precision evaluation is performed, and a first simulation regulation precision is output.

8. A virtual power plant operation regulation device adapted to multiple distributed resources, characterized by, The virtual power plant operation regulation method for adapting to the distributed resources comprises the following steps: a space construction module for constructing a power plant simulation space comprising a distributed resource layer, a monitoring and metering layer, and an operation regulation layer in a three-dimensional simulation platform; a sequence prediction module for monitoring and obtaining historical power grid operation data in a historical time zone through the monitoring and metering layer, predicting a predicted power load sequence, a predicted power grid frequency sequence, and a predicted renewable energy output ratio sequence in a preset time zone according to the historical power grid operation data, and determining a regulation response speed requirement; a scheme execution module for determining an adaptive resource combination strategy according to the regulation response speed requirement, performing operation regulation scheme optimization in the power plant simulation space based on the adaptive resource combination strategy with the goal of maximizing regulation precision, and determining an optimal regulation scheme for operation regulation.

Images