Virtual power plant optimal scheduling method considering renewable energy sources

Abstract

The invention discloses a virtual power plant optimal scheduling method considering renewable energy, and relates to the technical field of energy management, and the method comprises the steps: constructing a multi-energy cooperative control model, analyzing an operation feature data set, and dynamically adjusting the resource distribution of a virtual power plant; performing time period load alignment on the power in the power generation regulation and control scheme, and generating a user response strategy in combination with an intelligent demand response mechanism; performing combined scheduling on the power generation regulation and control scheme and the user response strategy, and outputting a power generation and utilization combined scheduling report; and the dispatching center executes power generation regulation and control and power utilization guidance on the virtual power plant according to the power generation and power utilization combined dispatching report, and feeds back and updates the multi-energy cooperative control model according to the operation data after execution and the dispatching target in the power generation and power utilization combined dispatching report. According to the invention, the multi-energy cooperative control model is constructed to analyze the operation characteristic data set, the resource distribution of the virtual power plant is dynamically adjusted, and the complementarity and coordination relationship among different energy forms are accurately identified.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of energy management, and in particular to a virtual power plant optimal scheduling method considering renewable energy. BACKGROUND

[0002] With the increasing penetration of renewable energy, virtual power plants, as an important means of aggregating distributed energy resources and improving grid regulation capabilities, play an increasingly critical role in modern power equipment. Traditional virtual power plant optimal scheduling methods usually construct feature variables based on historical operation data and use mathematical modeling methods such as mixed integer linear programming to coordinate the scheduling of multiple types of energy. Through load forecasting and demand response mechanisms, the generation and consumption sides are coordinated and controlled to form a complete scheduling scheme.

[0003] However, the existing methods still have certain limitations in dealing with complex and variable energy coordination, especially in the mapping accuracy between operation characteristics and scheduling decisions. On the one hand, conventional models often rely on static or semi-dynamic historical scheduling samples, which are difficult to accurately reflect the evolution of scheduling behavior caused by device state changes. On the other hand, there is a lack of fine-grained identification and dynamic adjustment mechanism for load mismatch periods, which affects the flexibility and adaptability of the overall scheduling. Existing technologies usually base on historical scheduling samples and static feature modeling to improve the mapping relationship between operation characteristics and scheduling decisions and the matching accuracy of user response strategies. SUMMARY

[0004] In view of the above existing problems, the present application is proposed.

[0005] Therefore, the present application provides a virtual power plant optimal scheduling method considering renewable energy to solve the problem of improving the mapping accuracy between operation characteristics and scheduling decisions.

[0006] To solve the above technical problems, the present application provides the following technical solutions: In a first aspect, the present application provides a virtual power plant optimal scheduling method considering renewable energy, which includes collecting and preprocessing the operation data of each distributed energy node in the virtual power plant to generate an operation feature dataset; constructing a multi-energy collaborative control model to analyze the operation feature dataset, dynamically adjusting the resource allocation of the virtual power plant, and obtaining a power generation regulation scheme; aligning the power in the power generation regulation scheme with the load by time period, and combining with an intelligent demand response mechanism to generate a user response strategy; jointly scheduling the power generation regulation scheme and the user response strategy to output a joint generation and consumption scheduling report; the scheduling center executes power generation regulation and power consumption guidance on the virtual power plant according to the joint generation and consumption scheduling report, generates a regulation signal, and feeds back and updates the multi-energy collaborative control model according to the operation data after execution and the scheduling target in the joint generation and consumption scheduling report.

[0007] As a preferred scheme of the virtual power plant optimal scheduling method considering renewable energy sources, wherein: the operation data includes power generation power, energy storage state, user load, environmental parameters and operation state of distributed energy node equipment; The preprocessing includes data cleaning, normalization processing, timestamp alignment and feature extraction; The preprocessed operation data is integrated to generate operation feature data set.

[0008] As a preferred scheme of the virtual power plant optimal scheduling method considering renewable energy sources, wherein: the operation feature data set is analyzed by constructing a multi-energy collaborative control model, and the specific steps are, Extract the operation feature vector in the operation feature data set, and extract the scheduling record corresponding to the operation feature vector in the historical scheduling result of the virtual power plant to generate a scheduling response variable; The operation feature vector and the scheduling response variable are divided into a collaborative scheduling training set, and modeling training is performed based on the collaborative scheduling training set to output a multi-energy collaborative control model; The operation feature data set is mapped into the multi-energy collaborative control model for field alignment and time synchronization to generate the mapped operation feature data.

[0009] As a preferred scheme of the virtual power plant optimal scheduling method considering renewable energy sources, wherein: the resource allocation of the virtual power plant is dynamically adjusted to obtain a power generation regulation scheme, and the specific steps are, Analyze the mapped operation feature data, identify the complementarity and coordination relationship between different energy forms, and generate an optimization input parameter; Based on the optimization input parameter, set the virtual power plant operation constraint condition in the multi-energy collaborative control model, set the cost function in the multi-energy collaborative control model, adjust the operation cost of the virtual power plant, and generate an optimization objective function; Use a standard optimization solver to solve the multi-energy collaborative control model containing the virtual power plant operation constraint condition and the optimization objective function, output the power generation regulation time series data set, and integrate it to output the power generation regulation scheme.

[0010] As a preferred scheme of the virtual power plant optimal scheduling method considering renewable energy sources, wherein: the power in the power generation regulation scheme is aligned with the period load, and the specific steps are, Obtain the user side historical load curve of the virtual power plant, and extract the total power supply power in the power generation regulation scheme; Superimpose the user side historical load curve and the total power supply power on a unified time axis, calculate the supply-demand difference in each time period as the period load alignment result.

[0011] As a preferred scheme of the virtual power plant optimal scheduling method considering renewable energy sources, wherein: the intelligent demand response mechanism is combined to generate a user response strategy, and the specific steps are, Obtaining the operation characteristics of the virtual power plant and the user electricity consumption behavior, and constructing an intelligent demand response mechanism; Identifying the load mismatch period in the period load alignment result, and combining the intelligent demand response mechanism to form a demand response regulation instruction; Integrating the period load alignment result and the demand response regulation instruction to generate a user response strategy.

[0012] As a preferred scheme of the virtual power plant optimal scheduling method considering renewable energy sources, wherein: the intelligent demand response mechanism is combined to generate a user response strategy, and the specific steps are, Performing hour-by-hour supply-demand matching analysis on the power generation regulation scheme and the user response strategy to generate a scheduling instruction set; Integrating the scheduling instruction set to output a generation and consumption joint scheduling report.

[0013] As a preferred scheme of the virtual power plant optimal scheduling method considering renewable energy sources, wherein: the scheduling center performs power generation regulation and electricity consumption guidance on the virtual power plant according to the generation and consumption joint scheduling report, generates a regulation signal, and feeds back and updates the multi-energy collaborative control model according to the operation data and the scheduling target in the generation and consumption joint scheduling report, and the specific steps are, Extracting the power generation regulation instruction and the electricity consumption guidance signal from the generation and consumption joint scheduling report, and issuing and executing them to the distributed energy and aggregated users in the virtual power plant to generate a regulation signal; Collecting the operation data of the virtual power plant after executing the regulation signal to generate post-execution operation data; Comparing the post-execution operation data with the scheduling target set in the generation and consumption joint report to generate a regulation execution result; Feeding back and updating the multi-energy collaborative control model according to the regulation execution result.

[0014] In a second aspect, the present application provides a computer device comprising a memory and a processor, wherein the memory stores a computer program, and wherein the computer program is executed by the processor to implement any step of the virtual power plant optimal scheduling method considering renewable energy sources according to the first aspect of the present application.

[0015] In a third aspect, the present application provides a computer readable storage medium having a computer program stored thereon, wherein the computer program is executed by a processor to implement any step of the virtual power plant optimal scheduling method considering renewable energy sources according to the first aspect of the present application.

[0016] The application has the beneficial effects that: by constructing a multi-energy collaborative control model to analyze the operation characteristic data set, dynamically adjusting the resource allocation of the virtual power plant, the complementary relationship and coordination relationship between different energy forms are accurately identified, and the optimized input parameters and power generation regulation scheme are generated. The power in the power generation regulation scheme is aligned with the time period load by the LSTM load prediction alignment method, and the user response strategy is generated by combining the intelligent demand response mechanism, so that the fine identification and dynamic adjustment of the user side load mismatch period are realized. BRIEF DESCRIPTION OF DRAWINGS

[0017] In order to more clearly illustrate the technical solutions of the embodiments of the present 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 present application, and other drawings can be obtained by those skilled in the art without creative labor.

[0018] Fig. 1 The flowchart of the virtual power plant optimization scheduling method considering renewable energy.

[0019] Fig. 2 The architecture diagram of the virtual power plant optimization scheduling.

[0020] Fig. 3 The flowchart of the multi-energy collaborative control model construction.

[0021] Fig. 4 The diagram of the LSTM load prediction alignment and response strategy generation. DETAILED DESCRIPTION

[0022] In order to make the above-mentioned purposes, features and advantages of the present application more apparent and easy to understand, the specific embodiments of the present application will be described in detail below with reference to the drawings of the specification.

[0023] In the following description, many specific details are set forth in order to provide a thorough understanding of the present application, but the present application can also be implemented in other ways different from those described herein, and those skilled in the art can make similar generalizations without departing from the connotation of the present application, therefore the present application is not limited to the specific embodiments disclosed below.

[0024] Secondly, the "one embodiment" or "embodiment" referred to herein means that the specific features, structures or characteristics can be included in at least one implementation of the present application. In this specification, "in one embodiment" does not mean the same embodiment, nor is it an independent or alternative embodiment that excludes other embodiments.

[0025] REFERENCE Figs. 1-4, is an embodiment of the present invention, which provides a virtual power plant optimization scheduling method considering renewable energy, including the following steps: S1. Collect and pre-process the operating data of each distributed energy node in the virtual power plant to generate an operating characteristic data set.

[0026] Operational data includes power generation, energy storage status, user load, environmental parameters, and the operating status of distributed energy node equipment; It should be noted that the generated power of wind farms and photovoltaic power stations is collected, and the actual output value of the generated power in different time periods is measured by smart meters, for example, once every 15 minutes, in kilowatts; the energy storage status of the energy storage equipment is collected, and the battery management equipment is used to record the state of charge percentage, current charge and discharge power, and battery temperature, for example, the state of charge is 65%, the charge and discharge power is 20 kilowatts, and the battery temperature is 30 degrees Celsius; the user load data is collected, and the actual power consumption of different users in different time periods is obtained with the help of user-side smart meters, for example, the user load in a certain period is 45 kilowatts; the environmental parameters are collected, and meteorological sensors are used to measure wind speed, light intensity and ambient temperature, for example, the wind speed is 6 meters per second, the light intensity is 800 watts per square meter, and the ambient temperature is 25 degrees Celsius; the operating status of distributed energy node equipment is collected, and the programmable logic controller is used to read whether the distributed energy node equipment is online, the fault status flag and the operating mode identification, for example, the equipment is in operation and there is no fault alarm.

[0027] Preprocessing includes data cleaning, normalization, timestamp alignment, and feature extraction; Specifically, data cleaning is first performed to identify and process outliers and missing values ​​in the collected data on power generation, energy storage status, user load, environmental parameters, and the operating status of distributed energy node devices. For example, power generation records exceeding the rated power range are eliminated, and missing energy storage status data are filled in using linear interpolation. Normalization is then performed to map the cleaned power generation, energy storage status, user load, and environmental parameter values ​​to a unified dimension range. For example, wind farm power generation is converted from kilowatts to a standardized value between 0 and 1 to facilitate comparison and calculation between data of different dimensions. Timestamp alignment is then performed to align time series data of power generation, energy storage status, user load, environmental parameters, and the operating status of distributed energy node devices from different sources to a unified time granularity. For example, all data is adjusted to a sampling point every 15 minutes, and time linear interpolation is used to fill in gaps caused by differences in sampling frequency. Finally, feature extraction is performed to extract statistical features and change trends from the aligned power generation, energy storage status, user load, and environmental parameters. For example, the sliding average of photovoltaic power generation or the diurnal fluctuation amplitude of user load is calculated.

[0028] The preprocessed operation data is integrated by a sliding window analysis method to generate an operation feature data set; Further, the preprocessed operation data is integrated in the time dimension by a sliding window analysis method, and the preprocessed operation data of continuous time periods is divided according to a set time window length and step, for example, 24 hours is used as the window length, and 1 hour is used as the step, and data segments of the power generation, energy storage state, user load, environmental parameters and operation state of the distributed energy node equipment in different time periods are sequentially intercepted; the time window length represents the time range covered by each analysis, which is used to intercept continuous preprocessed operation data, for example, the time window length is set to 24 hours, and the step represents the time interval of the window moving forward each time, which is used to control the overlap degree between adjacent windows, for example, the step is set to 1 hour, which means that after the data of one window is processed, the window is moved forward by 1 hour to continue to intercept the data of the next time period; The power generation, energy storage state, user load and environmental parameters in each window are counted, for example, mean, variance, maximum and minimum, and the operation state of the distributed energy node equipment is retained as a state identifier; the data segments and statistical features corresponding to all windows are arranged in time sequence to generate a structured operation feature data set.

[0029] S2, construct a multi-energy collaborative control model to analyze the operation feature data set, dynamically adjust the resource allocation of the virtual power plant, and obtain a power generation regulation scheme.

[0030] The operation feature vectors in the operation feature data set are extracted by using the time sequence feature method, and the scheduling records corresponding to the operation feature vectors in the historical scheduling results of the virtual power plant are extracted by using the response variable generation method to generate scheduling response variables; Specifically, the operation feature data set is segmented and sampled according to a fixed time granularity, for example, each hour is used as a sampling interval, and the data segments in each time period are sequentially intercepted; statistical features and trend features are calculated in each sampling interval, for example, mean, variance, change rate and moving average, to form a multi-dimensional numerical combination as the operation feature vector of each time period; the scheduling records in the corresponding time period are extracted from the historical scheduling results of the virtual power plant by using the response variable generation method, including the output adjustment instructions of the wind farm and photovoltaic power station, the charging and discharging arrangement of the energy storage device, and the user load adjustment strategy, for example, the output of the wind farm is reduced by 5% in a time period, the energy storage device enters the discharging state, and the user load is reduced by 30 kW; each scheduling record is converted into a numerical response variable, and finally the scheduling response variables corresponding to the operation feature vectors are generated.

[0031] The operation feature vector and the scheduling response variable are divided into a collaborative scheduling training set according to a sample division method, and modeling training is performed based on the collaborative scheduling training set by using a mixed integer linear programming method, and a multi-energy collaborative control model is output; Further, the operation feature vector and the scheduling response variable are divided into a data subset for training and a data subset for verification according to time sequence by using a sample division method, for example, 80% of the data is used for training and 20% is used for verification; ensure that each operation feature vector in the training set is time consistent with the corresponding scheduling response variable, and form a collaborative scheduling training set.

[0032] A mathematical optimization framework is constructed based on a mixed integer linear programming method, in which the operation feature vector is used as an input variable, the scheduling response variable is used as a target variable, and constraints such as power balance, charging and discharging capacity of energy storage devices, user load adjustment range, and operating state limits of distributed energy node devices are set, for example, the output of the wind farm does not exceed the maximum available power, the absolute value of the charging and discharging power of the energy storage device is not greater than the rated power, and the user load adjustment range does not exceed 20% of the baseline value; the maximum available power is usually determined based on the technical parameters of the device and the real-time environmental conditions, for example, the maximum available power of a wind turbine depends on the current wind speed, air density and rated capacity of the device; for a photovoltaic power station, it depends on the light intensity, temperature and efficiency of the photovoltaic module; the rated power refers to the power output or input capacity that a device should be able to achieve under normal working conditions according to standards. It is a fixed value defined by the device manufacturer according to the parameters and technical specifications of the device, which reflects the maximum carrying capacity of the device during long-term stable operation; the baseline value refers to the average level of a variable (such as user load) in a certain time period, which is used as a benchmark for evaluating the change range.

[0033] The objective function is solved on the collaborative scheduling training set to make the predicted scheduling response variable as close as possible to the actual scheduling record, for example, to minimize the weighted sum of output deviation and adjustment cost; the objective function is a mathematical expression for evaluating the pros and cons of the scheduling scheme, which is composed of multiple weighted items in the multi-energy collaborative control model, such as power deviation, adjustment cost and energy storage loss; by minimizing the objective function, the generated scheduling response variable is close to the actual record; and finally a multi-energy collaborative control model is output.

[0034] The operation feature data set is mapped into the multi-energy collaborative control model for field alignment and time synchronization by using a data mapping method, and a mapped operation feature data set is generated; It should be noted that the operation feature data set and the input fields required by the multi-energy collaborative control model are matched one by one according to the field naming rules, for example, the wind farm power is mapped to the corresponding wind farm output active power field in the multi-energy collaborative control model; The time series data corresponding to each field is subjected to time synchronization operation according to the time granularity set by the multi-energy collaborative control model, for example, all time series data are unified to a sampling point every 15 minutes, and linear interpolation method is used to process the time offset and missing of field data, and under the premise of known adjacent time point data, the approximate value of the missing time point is calculated through linear function to realize the alignment and integrity repair of time series; the set time granularity is usually determined according to the demand and scheduling period of the multi-energy collaborative control model, for example, shorter time granularity is used in the minute-level fast response scene, and longer time granularity can be used in day-ahead scheduling. Ensure that all running feature data is consistent with the input format of the multi-energy collaborative control model in time and field dimensions, and finally generate the mapped running feature data.

[0035] The mapped running feature data is analyzed using collaborative analysis method to identify the complementarity and coordination relationship between different energy forms, and the optimization input parameters are generated; Specifically, the mapped running feature data is subjected to multivariate correlation analysis, and the Pearson correlation coefficient is used to calculate the correlation strength between the energy outputs, for example, the negative correlation coefficient between wind power output and light intensity is-0.73; the output fluctuation law and response delay characteristics of various types of energy in different time periods are identified through time series cross-correlation analysis, for example, it is identified that the photovoltaic output is in high position from 10 am to 4 pm, and the wind power output is mainly concentrated in the night to the early morning; The Pearson correlation coefficient is used to calculate the correlation strength between the energy outputs, and the expression is, ; Among them, represents the correlation strength between the energy outputs; represents the index of a type of distributed energy node in the virtual power plant; represents the index of another type of energy node, which is used for correlation analysis with ; represents the running feature sequence of the energy node ; represents the running feature sequence of the energy node ; T represents the total number of time points; represents the sum of the product of and at the corresponding time point; represents the sum of over the entire time period; represents the sum of over the entire time period; The key parameters for characterizing the energy coordination characteristics are extracted based on the correlation strength between the energy outputs using a statistical analysis method, and the key parameters include a wind power photovoltaic complementary rate, a storage energy adjustment response time length, and a load and new energy output matching degree. The wind power photovoltaic complementary rate is calculated according to the power difference and superposition value of wind power and photovoltaic power in the same time period in historical output data. The storage energy adjustment response time length is derived from the charging and discharging records of the storage energy equipment, for example, the time interval from receiving an instruction to actually completing the execution is counted, and the average response time length is 5 minutes. The load and new energy output matching degree is calculated based on the overlapping area ratio of the user load curve and the new energy output curve. The user load curve refers to the record of the actual power consumption of the user side changing with time within a certain time range, which is used to reflect the time sequence characteristics of the power consumption behavior. The new energy output curve refers to the change trend of the actual output power of renewable energy equipment such as a wind farm within a certain time range, which reflects the volatility and intermittency characteristics of new energy generation. The key parameters are used as the optimization input parameters of the multi-energy coordination control model.

[0036] Based on the optimization input parameters, the virtual power plant operation constraint conditions are set in the multi-energy coordination control model using a constraint modeling method, and the cost function is set in the multi-energy coordination control model using a particle swarm optimization method to adjust the operation cost of the virtual power plant, thereby generating an optimization objective function. It should be noted that the optimization input parameters are used as the basic input of the constraint modeling, and the constraint modeling method is used to set the boundary of the generation power balance, the storage energy charging and discharging capacity, the user load adjustment range and the operation state of the distributed energy node equipment in the multi-energy coordination control model as the virtual power plant operation constraint conditions. For example, the total output power of the wind farm and the photovoltaic power station cannot be lower than the minimum power supply requirement, the charging and discharging power of the storage energy equipment cannot exceed the rated power, and the user load adjustment range cannot exceed 20% of the baseline value. The minimum power supply requirement refers to the total power generation of the wind farm and the photovoltaic power station during the operation of the virtual power plant, which must meet a minimum power supply level to ensure the basic stability and reliability of power supply. It is usually set based on load prediction or dispatching plan, for example, the virtual power plant needs to provide at least 10 megawatts of renewable energy output in a certain period of time to ensure the power demand of critical users. The rated power indicates that the charging of the storage energy equipment at any time cannot exceed the maximum value of the rated power, which is used to prevent overloading of the equipment and protect the battery life. For example, the rated charging and discharging power of a certain storage energy equipment is 500 kilowatts, so it can only charge or discharge 500 kilowatt-hours of energy per hour at most. The baseline value is usually a reference load level obtained from historical power consumption data. For example, the baseline load of a certain user in a certain period of time is 100 kilowatts, so the load after adjustment cannot be lower than 80 kilowatts or higher than 120 kilowatts.

[0037] The particle swarm optimization method is introduced to construct a cost function, and the wind farm power generation cost, photovoltaic power station power generation cost, energy storage device charging and discharging loss, and user side load adjustment incentive cost in the virtual power plant are included in the total cost objective function expression of the virtual power plant multi-energy collaborative scheduling, for example, the total operation cost is expressed in the form of weighted sum of unit kilowatt-hour generation cost and adjustment cost; the particle swarm optimization method is used to search for the optimal solution under the premise of meeting all operation constraints, adjust the weight distribution, minimize the operation cost, and finally generate the optimization objective function for regulating the virtual power plant operation strategy.

[0038] The multi-energy collaborative control model containing the virtual power plant operation constraints and the optimization objective function is solved using a standard optimization solver, and a power generation regulation time series data set is output, and the structured data integration method is used for integration, and a power generation regulation scheme is output. Further, the optimization objective function and the operation constraint condition are loaded as input to the standard optimization solver; the standard optimization solver minimizes the operation cost as the target under the premise of meeting all operation constraints, uses the interior point method to numerically solve the multi-energy collaborative control model, iteratively solves the constructed mixed integer linear programming problem through mathematical optimization algorithm under the premise of meeting the virtual power plant operation constraint condition, and outputs the wind farm and photovoltaic power station output adjustment value, energy storage device charging and discharging plan, and user side load adjustment instruction, which are integrated into a structured power generation regulation time series data set; the wind farm and photovoltaic power station output adjustment value refers to the specific power value determined by the optimal power generation regulation scheme obtained by solving the multi-energy collaborative control model in each scheduling period. The charging and discharging plan of the energy storage device refers to the charging or discharging state and the corresponding power arrangement of the energy storage device to be executed in each scheduling period. The user side load adjustment instruction refers to the power adjustment requirement issued to the user side in each scheduling period, which is used to participate in the supply and demand balance of power equipment.

[0039] The structured data integration method is used to classify and arrange the various types of regulation instructions in the power generation regulation time series data set according to time sequence, energy type and execution object, for example, the wind farm, photovoltaic power station, energy storage device and user side load adjustment instructions are organized by hour granularity, the energy type refers to different types of energy involved in power generation regulation, such as wind power, photovoltaic power, energy storage, etc., and the execution object refers to the specific equipment or user group acted on by the power generation regulation instruction, such as wind farm, photovoltaic power station and power user, and finally a complete power generation regulation scheme is output.

[0040] S3, align the power in the power generation regulation scheme with the period load, and generate a user response strategy combined with the intelligent demand response mechanism.

[0041] The user-side historical load curve of the virtual power plant is obtained by using a load synchronization method, and the total power supply power in the power generation regulation scheme is extracted by using a multi-source power time series superposition method. Specifically, the power consumption load records of various users in the virtual power plant are sorted according to a fixed time granularity, for example, taking every 15 minutes as a sampling point, and the user-side load data in the past 30 days is extracted to form a user-side historical load curve; the multi-source power time series superposition method is adopted to time-align the output adjustment value of the wind farm and the photovoltaic power station in the power generation regulation scheme and the charging and discharging plan of the energy storage device at each time point, and to perform numerical superposition at each time point, for example, summing the output of the wind farm of 10 megawatts, the output of the photovoltaic power station of 8 megawatts, and the discharging of the energy storage device of 2 megawatts in a certain period to obtain the total power supply power of 20 megawatts in the period, and the extraction of the total power supply power is completed.

[0042] The user-side historical load curve and the total power supply power are superimposed on a unified time axis by using the LSTM load prediction alignment method, and the supply-demand difference in each time period is calculated as the time period load alignment result; Further, the historical load data of the user side in the virtual power plant is collected and sorted according to a fixed time granularity to form time series data; the time series data is normalized and divided into a training set and a test set, for example, 80% of the data is used for training and 20% is used for verification; an LSTM neural network structure is constructed, including an input layer, a plurality of LSTM hidden layers and an output layer, the input data is the historical load sequence and related influencing factors such as temperature and date type, and the output is the load prediction value in the target time period; the network parameters are continuously adjusted by using the back propagation algorithm to minimize the prediction error on the training set, and the performance of the LSTM load prediction model is evaluated on the test set, the evaluation model performance is used to measure the prediction accuracy and stability of the LSTM load prediction model on the test data, the network parameters refer to the weights and biases used for data calculation in the LSTM load prediction model, which determine the influence mode of the load prediction value on the load prediction result, and finally an LSTM load prediction model is constructed; The LSTM load prediction model is used to perform time series prediction on the user-side historical load curve to generate a load prediction value with the same time granularity as the power generation regulation scheme, and the total power supply power is aligned to a unified time axis according to the same time granularity; on the unified time axis, the total power supply power and the corresponding load prediction value at each time point are compared item by item, and the supply-demand difference of each period is calculated, for example, the total power supply power is 20 megawatts and the load prediction value is 18 megawatts in a certain period, then the supply-demand difference in the period is 2 megawatts; finally, the supply-demand differences of all time periods are combined to form the time period load alignment result.

[0043] The operation characteristics and user power consumption behavior of the virtual power plant are obtained by using the load strategy engine method to construct an intelligent demand response mechanism. It should be noted that the power generation of the wind farm, the power generation of the photovoltaic power station, the charging and discharging state of the energy storage device, the user load and the environmental parameters of the virtual power plant are collected by using the load strategy engine method, and combined with the generation control scheme and the time period load alignment result, the typical operation mode under the coordinated control of multiple energy sources is identified; By analyzing the actual execution of the user side load adjustment instruction, the power distribution characteristics, load fluctuation and response behavior of various users in different time periods are extracted, and the adjustment capacity and response law are identified. For example, after receiving the price incentive signal, a certain type of commercial user can actively reduce the air conditioning load in the peak period, and the adjustment amplitude is about 10% to 20%. Based on the analysis result, combined with the load characteristics and response sensitivity of the user, the user response capacity classification rule is established, the user is divided into three categories of high response potential, medium response potential and low response potential, and the corresponding load adjustment mode is set according to different categories, such as guiding users to adjust independently based on price signals, so as to construct an intelligent demand response mechanism.

[0044] The load mismatch period in the time period load alignment result is identified by using the supply-demand difference threshold judgment method, and combined with the intelligent demand response mechanism, the demand response control instruction is formed; Specifically, the supply-demand difference of each time period in the time period load alignment result is compared with the set difference threshold, for example, the time period with the absolute value of supply-demand difference exceeding 5% of the total power supply is determined as the load mismatch period; According to the determination result, the load mismatch period with power supply shortage or excess is screened out. The set difference threshold is determined according to the tolerance deviation range of power supply and demand balance in the operation process of the virtual power plant, which is usually set by referring to historical load fluctuation characteristics, power supply stability requirements and control response capacity and other factors. For example, in the case of total power supply of 20 megawatts, the supply-demand difference threshold can be set to 5% to 10%; According to the preset user response capacity classification rule and adjustment mode in the intelligent demand response mechanism, the price signal based load reduction instruction is sent to the high response potential user in the load mismatch period, and the user is guided to increase power consumption in the power surplus period, and finally the differentiated demand response control instruction for different user groups is generated. The preset user response capacity classification rule and adjustment mode refers to selecting the appropriate user group and implementing the matching load adjustment measure in the load mismatch period according to the user response potential category and the corresponding control strategy. The corresponding control strategy is used to guide how to implement effective demand response control instruction to various users in the load mismatch period.

[0045] The time period load alignment result and the demand response control instruction are integrated by using the multi-dimensional rule mapping method to generate the user response strategy.

[0046] Further, according to the supply-demand difference value and the load mismatch type of each time point in the period load alignment result, it is determined whether the regulation action is triggered in each period; the demand response regulation instruction is aligned according to the unified time axis, so as to ensure that the regulation action is synchronized with the load mismatch period; the regulation action refers to the specific power consumption adjustment operation issued to the user in the load mismatch period according to the preset regulation strategy; the preset regulation strategy refers to the regulation rules and execution scheme formulated for different load mismatch types and user response ability classification in the intelligent demand response mechanism; On the energy type dimension, the influence of different energy forms such as wind power, photovoltaic and energy storage on user response behavior is distinguished, for example, when the wind and light output fluctuates greatly, the user is preferentially guided to adjust the flexible load; on the user dimension, combined with the user historical response record and the current regulation target, the specific load adjustment proportion and incentive amount are allocated, for example, the instruction of reducing 10% load is issued to a certain type of user and the incentive standard of 0.3 yuan per kilowatt is matched; finally, the information is structured and integrated through the multi-dimensional rule mapping method, and the user response strategy organized according to time, energy type and user category is formed.

[0047] S4, jointly scheduling the power generation regulation scheme and the user response strategy, and outputting a joint scheduling report.

[0048] The power generation regulation scheme and the user response strategy are matched and analyzed by the generation and consumption collaborative analysis method, and a scheduling instruction set is generated; Specifically, the time-period output adjustment value of the wind farm and the photovoltaic power station in the power generation regulation scheme, the charging and discharging plan of the energy storage device, and the load adjustment instruction of the user in the user response strategy are aligned on the unified time axis to generate the regulation action on the power generation side and the response action on the user side; in each time period, the total power supply and the user side expected load level are respectively summarized, and the supply-demand difference in the period load alignment result is matched and analyzed, for example, when the total power supply is 20 megawatts, the user side expected load is 18 megawatts, and there is 2 megawatts of power supply surplus, the period is in a state of supply greater than demand; According to the matching analysis result, the user response ability classification rule and the corresponding regulation strategy are comprehensively considered to determine whether the power generation output needs to be further adjusted and the user needs to be guided to increase power consumption; finally, the regulation action on the power generation side and the response action on the user side in each time period are combined to form a scheduling instruction set, The scheduling instruction set is integrated by using the scheduling summary method, and a joint scheduling report is output; It should be noted that the scheduling instruction set is summarized and arranged in chronological order to ensure that the hourly output adjustment value of the wind farm and the photovoltaic power station, the charging and discharging plan of the energy storage device, and the load adjustment instruction of the user are consistent in the time dimension; the scheduling instruction set of each period is classified and summarized according to the energy type and the execution object, for example, the total adjustment amount and the peak change of wind power, photovoltaic power, energy storage and different user groups in each period of the whole day are counted respectively; finally, it is summarized and organized in a structured form to form a complete scheduling summary including time distribution, energy type distribution and user response, which is used as a joint scheduling report of generation and consumption.

[0049] S5, the dispatching center executes power generation regulation and power consumption guidance on the virtual power plant according to the joint scheduling report of generation and consumption, generates a regulation signal, and feeds back and updates the multi-energy collaborative control model according to the running data after execution and the scheduling target in the joint scheduling report of generation and consumption.

[0050] The regulation instruction analysis method is used to extract the power generation regulation instruction and the power consumption guidance signal from the joint scheduling report of generation and consumption, and the regulation signal transmission method is used to issue and execute to the distributed energy and aggregated users in the virtual power plant, and the regulation signal is generated; Further, the regulation instruction analysis method is used to identify and extract the corresponding power generation regulation instruction in each time period in the joint scheduling report of generation and consumption, including the hourly output adjustment value of the wind farm and the photovoltaic power station, the charging and discharging plan of the energy storage device, and the power consumption guidance signal; the extracted regulation instructions are matched to the execution object categories in the virtual power plant in chronological order, including the corresponding distributed energy nodes and aggregated user groups; The regulation signal transmission method is used to send the power generation regulation instruction to the local control terminal of the wind farm, the photovoltaic power station and the energy storage device in the form of a communication protocol, and the power consumption guidance signal is issued to the user side energy management center; finally, after the instruction analysis and action response are completed at each execution end, the regulation signal is generated.

[0051] The running data of the virtual power plant after executing the regulation signal is collected to generate the running data after execution; Specifically, after the regulation signal is issued and executed, the actual output value of the wind farm and the photovoltaic power station, the charging and discharging power of the energy storage device, and the actual load change of the aggregated users are collected in real time according to the unified time granularity, and the running state records of each distributed energy node and user group in each time period are collected, for example, the actual output of the wind farm in a certain period is 18 megawatts, the energy storage device discharges 3 megawatts, and the load of a certain aggregated user group decreases by 2.5 megawatts; the collected running state records are classified and arranged to form the running data after execution.

[0052] The post-execution operation data is compared with the scheduling target set in the joint report of generation and consumption by a quantitative evaluation method, to generate a regulation and control execution result, and the multi-energy collaborative control model is fed back and updated according to the regulation and control execution result.

[0053] It should be noted that the quantitative evaluation method is used to compare the post-execution operation data with the scheduling target values in the joint scheduling report of generation and consumption in the corresponding time period, to obtain the execution deviation of each regulation and control action, for example, the target output of a wind farm in a certain period is 20 MW, and the actual output is 19.2 MW, and the deviation is 0.8 MW; the regulation and control action refers to the specific operation of adjusting the output of the power generation equipment in a specific time period according to the scheduling instruction; Based on the execution deviations of all time points, the key performance indicators such as the average absolute error, the maximum deviation period and the response delay time are counted to form the regulation and control execution result; the regulation and control execution result is input into the multi-energy collaborative control model as feedback, to identify the prediction deviation of the multi-energy collaborative control model in the optimization solving process, and to adjust the parameters of the multi-energy collaborative control model, to complete the iterative update of the multi-energy collaborative control model.

[0054] The embodiment also provides a computer device suitable for the case of the virtual power plant optimal scheduling method considering renewable energy, which comprises a memory and a processor; the memory is used to store computer executable instructions, and the processor is used to execute the computer executable instructions to realize the virtual power plant optimal scheduling method considering renewable energy proposed in the above embodiment.

[0055] The computer device can be a terminal, and the computer device comprises a processor, a memory, a communication interface, a display screen and an input device connected through a system bus. The processor of the computer device is used to provide computing and control capabilities. The memory of the computer device comprises a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system and a computer program. The internal memory provides an environment for the operation of the operating system and the computer program in the non-volatile storage medium. The communication interface of the computer device is used to communicate with external terminals in a wired or wireless manner. The wireless manner can be realized through WIFI, an operator network, NFC (near field communication) or other technologies. The display screen of the computer device can be a liquid crystal display screen or an electronic ink display screen. The input device of the computer device can be a touch layer overlaid on the display screen, or a key, trackball or touchpad arranged on the shell of the computer device, or an external keyboard, touchpad or mouse, etc.

[0056] The embodiment also provides a storage medium on which a computer program is stored, the program being executed by a processor to implement the method for optimizing scheduling of a virtual power plant considering renewable energy as proposed in the above embodiment; the storage medium can be implemented by any type of volatile or non-volatile storage device or a combination thereof, such as a static random access memory (SRAM), an electrically erasable programmable read-only memory (EEPROM), an erasable programmable read-only memory (EPROM), a programmable read-only memory (PROM), a read-only memory (ROM), a magnetic memory, a flash memory, a magnetic disk, or an optical disk.

[0057] To sum up, the present application realizes accurate identification of the complementarity and coordination relationship between different energy forms by constructing a multi-energy collaborative control model to analyze the operation characteristic data set and dynamically adjusting the resource allocation of the virtual power plant, and generates an optimized input parameter and a power generation regulation scheme. The power in the power generation regulation scheme is aligned with the time period load by using the LSTM load prediction alignment method, and a user response strategy is generated in combination with an intelligent demand response mechanism, thereby realizing fine identification and dynamic adjustment of the user side load mismatch period.

[0058] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present application but not limit the present application. Although the present application has been described in detail with reference to the preferred embodiments, those skilled in the art should understand that the technical solutions of the present application can be modified or replaced equivalently without departing from the spirit and scope of the technical solutions of the present application, and all of them should be covered in the scope of the claims of the present application.

Claims

1. A virtual power plant optimization scheduling method considering renewable energy, characterized by: include, Collect and pre-process the operating data of each distributed energy node in the virtual power plant to generate an operating characteristic data set; Build a multi-energy collaborative control model to analyze operational characteristic data sets, dynamically adjust resource allocation in virtual power plants, and obtain power generation control solutions; Align the power in the power generation control plan with the load during each period, and combine it with the intelligent demand response mechanism to generate a user response strategy; Combine power generation control schemes and user response strategies to jointly dispatch and output a combined generation and consumption dispatch report; The dispatching center performs power generation control and power consumption guidance on the virtual power plant based on the joint dispatching report of power generation and consumption, generates control signals, and provides feedback and updates to the multi-energy collaborative control model based on the post-execution operation data and the dispatching targets in the joint dispatching report of power generation and consumption.

2. The virtual power plant optimization scheduling method considering renewable energy as claimed in claim 1, characterized in that: The operating data includes power generation, energy storage status, user load, environmental parameters and operating status of distributed energy node equipment; The preprocessing includes data cleaning, normalization, timestamp alignment and feature extraction; The preprocessed operation data is integrated to generate an operation feature dataset.

3. The virtual power plant optimization scheduling method considering renewable energy as claimed in claim 1, characterized in that: The specific steps of constructing a multi-energy coordinated control model to analyze the operation characteristic data set are: Extract the operation feature vector from the operation feature data set, and extract the scheduling record corresponding to the operation feature vector from the historical scheduling results of the virtual power plant to generate the scheduling response variable; The operating characteristic vectors and dispatch response variables are divided into a collaborative dispatch training set, and modeling training is performed based on the collaborative dispatch training set to output a multi-energy collaborative control model; The operation characteristic data set is mapped to the multi-energy collaborative control model, and field alignment and time synchronization are performed to generate mapped operation characteristic data.

4. The virtual power plant optimization scheduling method considering renewable energy as claimed in claim 3, characterized in that: The dynamic adjustment of the resource allocation of the virtual power plant and the acquisition of the power generation control plan are specifically carried out as follows: Analyze the mapped operational characteristic data, identify the complementarity and coordination relationships between different energy forms, and generate optimized input parameters; Based on the optimized input parameters, the operating constraints of the virtual power plant are set in the multi-energy collaborative control model, and the cost function is set in the multi-energy collaborative control model to adjust the operating cost of the virtual power plant and generate the optimization objective function; A standard optimization solver is used to solve the multi-energy collaborative control model that includes the virtual power plant operation constraints and optimization objective functions, output the power generation control time series data set, integrate it, and output the power generation control plan.

5. The virtual power plant optimization scheduling method considering renewable energy as claimed in claim 1, characterized in that: The specific steps of aligning the power load in the power generation control scheme are as follows: Obtain the user-side historical load curve of the virtual power plant and extract the total power supply power in the power generation control plan; The user-side historical load curve and total power supply are superimposed on a unified time axis, and the supply-demand difference in each time period is calculated as the time period load alignment result.

6. The virtual power plant optimization scheduling method considering renewable energy as claimed in claim 5, characterized in that: The above method combines the intelligent demand response mechanism to generate a user response strategy. The specific steps are: Obtain the operating characteristics of virtual power plants and user electricity consumption behaviors to build an intelligent demand response mechanism; Identify the load mismatch period in the load alignment results, and combine it with the intelligent demand response mechanism to form demand response control instructions; Integrate the time period load alignment results and demand response control instructions to generate a user response strategy.

7. The virtual power plant optimization scheduling method considering renewable energy as claimed in claim 6, characterized in that: The specific steps of jointly scheduling the power generation control plan and the user response strategy and outputting the power generation and user joint scheduling report are as follows: Perform supply and demand matching analysis on each time period between the power generation control plan and the user response strategy to generate a dispatch instruction set; Integrate the scheduling instruction set and output the joint scheduling report.

8. The virtual power plant optimization scheduling method considering renewable energy as claimed in claim 1, characterized in that: The dispatching center performs power generation control and power consumption guidance on the virtual power plant according to the joint dispatching report of power generation and consumption, generates a control signal, and provides feedback and updates to the multi-energy collaborative control model based on the post-execution operation data and the dispatching targets in the joint dispatching report of power generation and consumption. The specific steps are as follows: Extract power generation control instructions and power consumption guidance signals from the joint dispatch report of power generation and consumption, and issue and execute them to distributed energy resources and aggregated users in the virtual power plant to generate control signals; Collect the operating data of the virtual power plant after executing the control signal and generate post-execution operating data; Compare the post-execution operation data with the scheduling targets set in the joint report of generation and use to generate the control execution results; Feedback and update of the multi-energy collaborative control model are carried out according to the control execution results.

9. A computer device comprising a memory and a processor, wherein the memory stores a computer program, wherein: When the processor executes the computer program, the steps of the virtual power plant optimization scheduling method considering renewable energy are implemented as described in any one of claims 1 to 8.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the steps of the virtual power plant optimization scheduling method considering renewable energy are implemented as described in any one of claims 1 to 8.

Images