Multi-time scale scheduling method and system based on virtual power plant

Abstract

The invention provides a multi-time scale scheduling method and system based on a virtual power plant. According to the implementation scheme, a day-ahead objective function is constructed by taking benefit maximization of a day-ahead power dispatching scheme of a virtual power plant on the next day as a target and taking response of each power device in the virtual power plant to a first type of load as a condition; optimizing to obtain a day-ahead scheduling curve of each gas turbine unit, each wind turbine unit, each photovoltaic unit, each energy storage device and each first type of load on the next day; performing power dispatching on the virtual power plant in the next day based on the day-ahead dispatching curve, and monitoring actual output; under the condition that the actual output of a wind turbine generator and / or a photovoltaic unit in a first time period in the next day is not matched with the predicted output, the adjustment cost of an intra-day power dispatching scheme of a virtual power plant in the first time period is minimized, and the condition that a gas turbine generator in the virtual power plant responds to a second type of load is taken as the condition; and constructing an intra-day target function to optimize and obtain an intra-day scheduling curve of the gas unit and the first type of load in the first time period. Therefore, the safe and stable operation of the power system can be ensured.

Description

Technical Field

[0001] The present invention relates to the field of electric power technology, and in particular to a multi-time scale scheduling method and system based on a virtual power plant. Background Art

[0002] With the continuous development of user-side resources and advancements in energy storage technology, user-side resources will no longer be limited to the role of consumers but will gradually shift to prosumers. Consequently, user-side resources will no longer be limited to load resources but will also expand to include energy storage equipment and distributed generators. However, distributed generators and flexible load resources have characteristics such as small capacity, relatively dispersed geographical distribution, and numerous influencing factors. Therefore, how to manage user-side resources and maximize their effectiveness has become a key and difficult issue in the transformation and upgrading of the energy and power industry in the new era.

[0003] As a business model and mode that can break through geographical restrictions and widely gather diverse user-side resources such as multiple heterogeneous energy users, distributed resources, and energy storage, virtual power plants can, on the one hand, promote the consumption of renewable energy and ensure the safe and stable operation of the power grid to a certain extent. On the other hand, in the "Internet +" environment, the deepening integration of digital technology and the technical framework of the energy industry provides new ideas and technical solutions for improving the operating efficiency of virtual power plants. Virtual power plants can effectively manage decentralized user-side resources based on advanced measurement, communication, and control technologies, and thus improve their overall benefits through reasonable coordination and regulation while meeting internal resource needs. In summary, relying on the "electricity + computing power" combined architecture system to accelerate the construction and upgrading of virtual power plants and promote the transformation of user-side resources to networking, intelligence, and digitalization has become one of the key directions for the construction of new power systems. Summary of the Invention

[0004] The present invention provides a multi-time scale scheduling method and system based on a virtual power plant, which can solve at least one of the above technical problems.

[0005] According to one aspect of the present invention, a multi-time-scale scheduling method based on a virtual power plant is provided, comprising:

[0006] A day-ahead objective function is constructed with the goal of maximizing the revenue of the virtual power plant's next-day day-ahead power dispatch plan and with the condition that each power device in the virtual power plant responds to the first type of load and does not respond to the second type of load;

[0007] Based on the day-ahead dispatch constraints of the virtual power plant, solving the day-ahead objective function for a maximum value to obtain a day-ahead dispatch curve for each of the power equipment and the first type of load on the next day, wherein the power equipment includes a gas generator set, a wind generator set, a photovoltaic generator set, and an energy storage device;

[0008] Based on the day-ahead dispatch curve, the virtual power plant is dispatched for power on the next day, and the actual output of the wind turbine and photovoltaic unit is monitored in each time period on the next day;

[0009] In a case where the actual output of the wind turbine generator set and / or the photovoltaic generator set in the first time period of the next day does not match the predicted output for the corresponding time period in the corresponding day-ahead dispatch curve, an intraday objective function is constructed with the goal of minimizing the adjustment cost of the intraday power dispatch plan of the virtual power plant in the first time period and with the condition that the gas generator set in the virtual power plant responds to the second type of load;

[0010] Based on the intraday scheduling constraints of the virtual power plant, the minimum value of the intraday objective function is solved to obtain the intraday scheduling curve of the gas unit and the first type of load in the first time period.

[0011] According to another aspect of the present invention, a multi-time-scale scheduling device based on a virtual power plant is provided, comprising:

[0012] The first function construction module is configured to construct a day-ahead objective function with the goal of maximizing the revenue of the virtual power plant's day-ahead power dispatch plan on the next day and with the condition that each power device in the virtual power plant responds to the first type of load and does not respond to the second type of load;

[0013] a first optimization module, configured to solve the day-ahead objective function for a maximum value based on the day-ahead dispatch constraints of the virtual power plant, and obtain a day-ahead dispatch curve for each of the power equipment and the first type of load on the next day, wherein the power equipment includes a gas generator set, a wind turbine set, a photovoltaic generator set, and an energy storage device;

[0014] a first scheduling module, configured to perform power scheduling for the virtual power plant on the next day based on a day-ahead scheduling curve, and monitor actual outputs of the wind turbine generator set and the photovoltaic generator set in various time periods on the next day;

[0015] a second function construction module for constructing an intraday objective function with the goal of minimizing the adjustment cost of the intraday power dispatch plan of the virtual power plant in the first time period, and with the gas-fired units in the virtual power plant responding to the second type of load as a condition, when the actual output results of the wind turbine generator set and / or the photovoltaic generator set in the first time period of the next day do not match the predicted output for the corresponding time period in the corresponding day-ahead dispatch curve;

[0016] The second optimization module is used to solve the minimum value of the intraday objective function based on the intraday scheduling constraints of the virtual power plant, and obtain the intraday scheduling curve of the gas unit and the first type of load in the first time period.

[0017] According to the technical solution of the present invention, two types of loads are used to respectively schedule the virtual power plant on a day-ahead and intraday basis. First, the first type of load is used to optimize the day-ahead scheduling plan of the virtual power plant with the goal of maximizing revenue, and the day-ahead scheduling curve of the next day is obtained. When power scheduling is performed according to the curve on the next day, if it is monitored that the actual wind power output and / or actual photovoltaic output of a certain time period of the next day does not match the corresponding predicted output in the curve, the second type of load of the time period is used, with the goal of minimizing the output adjustment cost, to optimize the intraday scheduling plan of the virtual power plant in the time period, and obtain the intraday scheduling curve of the time period. Thus, scheduling is performed for the first time period according to the intraday scheduling curve. In this way, different loads and outputs are used for scheduling at different time scales of power scheduling on the day-ahead and intraday basis, and internal resources can be quickly adjusted to correct the deviation of the day-ahead scheduling plan, thereby ensuring that the power system can operate safely and stably.

[0018] It should be understood that the content described in this section is not intended to identify the key or important features of the embodiments of the present invention, nor is it intended to limit the scope of the present invention. Other features of the present invention will become readily understood through the following description. BRIEF DESCRIPTION OF THE DRAWINGS

[0019] The accompanying drawings are provided for a better understanding of the present invention and do not constitute a limitation of the present invention.

[0020] Figure 1 is a flow chart of a multi-time-scale scheduling method based on a virtual power plant according to an embodiment of the present invention;

[0021] Figure 2 Schematic diagram of day-ahead scheduling and intraday scheduling according to an embodiment of the present invention;

[0022] Figure 3 is a flow chart of a training process of a wind power output prediction model according to an embodiment of the present invention;

[0023] Figure 4 This is a structural block diagram of a multi-time-scale scheduling device based on a virtual power plant according to an embodiment of the present invention;

[0024] Figure 5 is a block diagram of an electronic device for implementing the method according to an embodiment of the present invention. DETAILED DESCRIPTION

[0025] The following description of exemplary embodiments of the present invention is made in conjunction with the accompanying drawings, and various details of the embodiments of the present invention are included to facilitate understanding. These details should be considered as merely exemplary. Therefore, it should be appreciated by those skilled in the art that various changes and modifications may be made to the embodiments described herein without departing from the scope of the present invention. Similarly, for the sake of clarity and conciseness, descriptions of well-known functions and structures are omitted in the following description.

[0026] Figure 1 This is a flowchart of a multi-time-scale scheduling method based on a virtual power plant according to an embodiment of the present invention.

[0027] like Figure 1 As shown, the multi-time scale scheduling method based on virtual power plant may include:

[0028] S110, constructing a day-ahead objective function with the goal of maximizing the revenue of the virtual power plant's day-ahead power dispatch plan on the next day and with the condition that each power device in the virtual power plant responds to the first type of load and does not respond to the second type of load;

[0029] S120 , based on the day-ahead dispatch constraints of the virtual power plant, solving for the maximum value of the day-ahead objective function to obtain the day-ahead dispatch curves for each power device and the first type of load for the next day, where the power devices include gas turbines, wind turbines, photovoltaic units, and energy storage devices;

[0030] S130, based on the day-ahead dispatch curve, dispatching power to the virtual power plant on the next day, and monitoring the actual output of the wind turbines and photovoltaic units in various time periods on the next day;

[0031] S140, when the actual output of the wind turbine and / or photovoltaic unit in the first time period of the next day does not match the predicted output of the corresponding time period in the corresponding day-ahead dispatch curve, constructing an intraday objective function with the goal of minimizing the adjustment cost of the intraday power dispatch plan of the virtual power plant in the first time period and with the gas units in the virtual power plant responding to the second type of load as a condition;

[0032] S150, based on the intraday dispatch constraints of the virtual power plant, solving the minimum value of the intraday objective function to obtain the intraday dispatch curve of the gas unit and the first type of load in the first time period.

[0033] Exemplarily, the various power equipment of the virtual power plant responds to the first type of load, including: gas turbines, wind turbines, photovoltaic turbines and energy storage equipment, which respectively generate electricity or charge and discharge electricity for the first type of load, and the power grid that interacts with the virtual power plant purchases or sells electricity.

[0034] For example, the first type of load (Type I load) is the load that is notified one day in advance before the day-ahead scheduling, and the second type of load (Type II load) is the load that is notified one time in advance before the day-ahead scheduling.

[0035] Understandably, due to the highly planned and predictable nature of Class I loads, which tend to respond more slowly, the virtual power plant operator notifies users of adjustments for each time period the following day based on the day's optimized operation results. Since Class II loads are highly responsive and can be disconnected promptly, the virtual power plant promptly notifies users of the optimized operation plan every 15 minutes throughout the day.

[0036] For example, the day-ahead dispatch curve of the power equipment may include the predicted output of the power equipment in each time period.

[0037] like Figure 2 As shown in the figure, on the day-ahead timescale, with the goal of maximizing revenue and minimizing costs, Class I flexible loads are selected to participate in demand response. While satisfying various resource constraints and the power balance, the virtual power plant is optimized for operation. This ultimately results in a 24-hour day-ahead dispatch curve for the next day, encompassing gas turbines, distributed wind turbines, distributed photovoltaic units, energy storage equipment, Class I loads, and the interaction with grid power. On the intraday timescale, when wind and solar output deviates, based on the day-ahead optimization results and ultra-short-term renewable energy forecasts, the virtual power plant aims to maximize operating revenue (i.e., minimize operating costs). While satisfying various resource constraints and the power balance, the deviation is corrected by disconnecting Class II loads, rationally adjusting gas turbine output, and temporarily purchasing electricity from the grid. This ultimately results in adjustments to Class II loads, gas turbine output, and grid power purchases. This example can thus mitigate overall output fluctuations caused by fluctuations in renewable energy output, namely wind and photovoltaic power, while fully utilizing load resources and ensuring safe, stable, and economical operation of the virtual power plant.

[0038] According to the above implementation, two types of loads are used to perform day-ahead and intraday scheduling of the virtual power plant, respectively. First, the first type of load is used to optimize the day-ahead scheduling plan of the virtual power plant with the goal of maximizing revenue, and the day-ahead scheduling curve of the next day is obtained. When power scheduling is performed according to the curve the next day, if it is monitored that the actual wind power output and / or actual photovoltaic output in a certain time period of the next day does not match the corresponding predicted output in the curve, the second type of load in the time period is used, with the goal of minimizing the output adjustment cost, to optimize the intraday scheduling plan of the virtual power plant in the time period, and obtain the intraday scheduling curve of the time period. Thus, scheduling is performed for the first time period according to the intraday scheduling curve. In this way, different loads and outputs are used for scheduling at different time scales of power scheduling on the day-ahead and intraday basis, and internal resources can be quickly adjusted to correct the deviation of the day-ahead scheduling plan, thereby ensuring that the power system can operate safely and stably.

[0039] In one embodiment, the first type of load is the load that is notified one day in advance to participate in the regulation, and the electricity price of the first type of load is determined based on the price coefficient corresponding to the load change rate; the second type of load is the load that is notified a time period in advance to participate in the regulation, and the electricity price of the second type of load is determined based on the load adjustment power of the second type of load.

[0040] For example, Class I loads can undergo price-based demand response. During the day-ahead dispatch optimization process, changes in time-of-use electricity prices can be used to guide users to adjust their electricity usage, reducing their consumption during peak hours and increasing it during off-peak hours. The relationship between load changes and electricity price changes can be expressed using a price elasticity coefficient.

[0041] For example, the price elasticity coefficient can be expressed using the following formula:

[0042]

[0043] Among them, η load,rv is the price elasticity coefficient; ΔQ r is the load power change at time t; Δp r is the load price variable at time t; Q r is the load power at time t; p r is the load electricity price at time t; when r=v, η load is the elastic coefficient; when r≠v, η load is the cross elastic coefficient.

[0044] For example, based on the elastic price coefficient, the following elastic price matrix for the first type of load can be obtained:

[0045]

[0046] For example, Class II loads have a shorter daily adjustment time. Selecting Class II loads that can respond quickly to participate in demand response is one way to correct for deviations. When renewable energy output deviations or emergencies occur, the virtual power plant operator directly issues signals to Class II load users, requesting adjustments based on the actual situation. After the adjustments are made, settlement is made based on the fixed subsidy agreed in the pre-contract. If the user refuses to participate in the response, the virtual power plant operator will penalize the user in accordance with the contract.

[0047] For example, the calculation formula for the adjustable power of Class II load is as follows:

[0048]

[0049] in, is the adjustable power of Class II load; ωII,t Takes 0 or 1, indicating whether to adjust at time t, k t is the reduction coefficient, It is the power before adjustment of Class II load.

[0050] According to the above implementation, two different loads are used to optimize the day-ahead dispatching plan and the intraday dispatching plan of the virtual power plant respectively. Internal resources can be quickly adjusted to correct the deviation of the day-ahead dispatching plan, ensuring that the power system can operate safely and stably.

[0051] In one embodiment, a day-ahead objective function is constructed with the goal of maximizing the revenue of the virtual power plant's next-day day-ahead power dispatch plan and with the condition that each power device in the virtual power plant responds to the first type of load and does not respond to the second type of load, including:

[0052] Based on the upward and downward power regulation of the first type of load at various times in the next day, and the electricity sales price of the virtual power plant at various times in the next day, determine the revenue function of the virtual power plant in selling electricity to the internal load on the next day;

[0053] The day-ahead objective function is determined based on the revenue function, the virtual power plant's gas-fired power generation cost, wind power generation cost, photovoltaic power generation cost, solar curtailment penalty cost, wind curtailment penalty cost, and energy storage charging and discharging cost, as well as the electric energy interaction cost between the virtual power plant and the power grid.

[0054] For example, the profit function is:

[0055]

[0056] in, represents the total revenue of the virtual power plant from selling electricity to internal loads on the next day, represents the upward adjustment power of the first type of load at time t, represents the downward adjustment power of the first type of load at time t, D t represents the unregulated power of all loads in the virtual power plant, p l,t represents the time-of-use electricity price of the virtual power plant supplying power to the internal load at time t, Δp l,t It represents the changing price of the time-of-use electricity price when the virtual power plant supplies power to the internal load at time t.

[0057] For example, the calculation formula of the unadjusted power of all loads in the virtual power plant is as follows:

[0058]

[0059] in, represents the unadjusted power of the first type of load at time t, represents the unadjusted power of the second type of load at time t, Represents the rigid load in the virtual power plant.

[0060] For example, the day-ahead objective function can be obtained by subtracting the virtual power plant's gas-fired power generation cost, wind power generation cost, photovoltaic power generation cost, solar curtailment penalty cost, wind curtailment penalty cost, and energy storage charging and discharging cost from the revenue function, as well as the electric energy interaction cost between the virtual power plant and the power grid.

[0061] For example, the gas power generation cost of a gas-fired unit is:

[0062]

[0063] Among them, P gas,t is the output power of the gas generator set at time t; P gas,t is the marginal cost of the gas unit power output at time t, and Δt1 is the unit time, i.e. 1h.

[0064] For example, the cost of wind power generation by a wind turbine is:

[0065]

[0066] in, The distributed wind power dispatch output is optimized for the day before; c w The marginal cost of generating electricity from distributed wind turbines.

[0067] For example, the photovoltaic power generation cost of the photovoltaic power generation group is:

[0068]

[0069] in, Distributed photovoltaic dispatch output optimized for the day before; c pv The marginal cost of power generation from distributed photovoltaic units.

[0070] For example, the sum of the penalty costs for curtailing solar power and wind power in a virtual power plant is:

[0071]

[0072] Where, P aw,t is the wind power abandoned at time t; P apv,t is the abandoned optical power at time t; P abondon The penalty price for abandoning solar or wind power.

[0073] For example, the electric energy interaction cost between the virtual power plant and the power grid is:

[0074]

[0075] in, The power of the virtual power plant buying and selling electricity to the grid at time t; The price at which the virtual power plant buys and sells electricity to the main power grid at time t.

[0076] For example, the energy storage charging and discharging cost is:

[0077]

[0078] Among them, P ch,t 、P dis,t is the charging and discharging power of the energy storage at time t; p ch,t 、p dis,t is the cost of charging and discharging the energy storage at time t.

[0079] According to the above implementation mode, by adopting the above day-ahead objective function, the virtual power plant can optimize its operation with the goal of maximizing profit, and obtain the day-ahead dispatch curve of each resource, including the 24-hour gas unit output curve divided by 1 hour, the wind and solar output curve, the energy storage charging and discharging status, the interaction with the large power grid, and the demand response status of Class I loads.

[0080] In one embodiment, the intraday objective function is:

[0081]

[0082] Where C represents the intraday objective function, ΔC gas represents the cost of adjusting the gas-fired power generation output of the gas-fired unit, ΔC grid It represents the cost of the virtual power plant temporarily buying electricity from the grid. represents the adjustment cost of the second type of load, ΔP gas,t represents the adjusted power of the gas generator set at time t, Δp gas,t represents the marginal cost of output adjustment of the gas unit at time t, represents the power that the virtual power plant buys from the grid at time t; represents the price of electricity purchased by the virtual power plant from the grid at time t, Indicates the adjusted power of the second type of load, represents the subsidy cost of the second type of load, and Δt2 represents the duration of the first time period.

[0083] In this example, since there may be errors in the renewable energy output forecast, in order to ensure the safe and stable operation of the virtual power plant, based on the deviation between the intraday ultra-short-term renewable energy forecast and the day-ahead forecast, the virtual power plant uses the above-mentioned intraday objective function to adjust the deviation in three forms: adjusting the output of the gas turbine unit, temporarily purchasing electricity from the power grid, and Class II load participating in demand response.

[0084] The above embodiment can be applied to the cloud-edge-device collaboration framework. An example of the framework is introduced below, as follows:

[0085] The cloud-edge-end collaboration framework consists of three parts: the cloud center layer, the edge layer, and the terminal layer. The cloud center layer has advanced computing modules and information interaction systems that can realize functions such as analysis, prediction, and decision-making. While receiving and collecting various information and data from the edge layer, it also transmits instructions and signals to the edge layer. The edge side serves as a buffer zone between the cloud center layer and the terminal layer. Its main functions include storing information, preliminary calculations, and decomposing commands. When the cloud center layer issues a command, the edge layer decomposes it into sub-commands and then transmits it to the terminal layer. Similarly, when the terminal layer transmits information upward, the edge layer performs preliminary integrated calculations, one part of which is stored in the edge layer, and the other part is uploaded to the cloud center layer. The intelligent terminal layer is mainly responsible for collecting data, information, and monitoring equipment, transmitting standardized data and information to the edge layer as required, and accepting instructions from the cloud center layer and the edge layer.

[0086] The cloud-edge-device collaboration framework can provide new ideas and solutions to problems encountered during the operation of virtual power plants:

[0087] First, virtual power plants generate a vast amount of data and information that needs to be processed while managing and coordinating a large number of heterogeneous resources. Leveraging a cloud-edge-end collaborative framework, the terminal layer can standardize data and information transmission, enabling orderly collection and precise metering. This large amount of data can be placed at different levels and nodes for calculation and storage according to established rules, enabling virtual power plants to efficiently organize, filter, and utilize this data and information.

[0088] Secondly, the internal resources of virtual power plants are relatively dispersed, and physical isolation makes the efficient transmission and protection of critical data and information extremely crucial. Virtual power plants have extremely high real-time requirements during coordination and control. However, due to the traditional centralized computing model, all data accumulates in the cloud center awaiting calculation, which greatly complicates the management and utilization of data and information, and also poses significant risks to security and privacy. The cloud-edge-end collaborative framework, with its high-speed transmission of data and information and the interconnected analysis and processing of each node, enables efficient control of virtual power plants at the functional level.

[0089] Finally, the large-scale access of renewable energy across regions and the demand for diversified load resources make the virtual power plant extremely uncertain and complex. This requires relying on the cloud-edge-end collaborative framework to comprehensively control, analyze and predict the internal resources of the virtual power plant, coordinate and regulate from a global perspective, and realize the safe and stable operation of the virtual power plant.

[0090] Based on the above analysis, the organic combination of the cloud-edge-end collaborative functional framework and the virtual power plant can realize the management, transmission and utilization of data information such as information flow, energy flow, and business flow.

[0091] The operation mode of the virtual power plant in the above embodiment may be as follows:

[0092] The virtual power plant is divided into two layers, the upper layer is the virtual power plant operator; the lower layer is the user-side resources, including distributed power generation equipment, energy storage equipment, third-party load integrators, large users, and industrial parks.

[0093] 1. At the upper level, the virtual power plant operator manages internal resources like the human brain and responds to external influences. The virtual power plant operator's goals are, first, to ensure a balance between supply and demand, meet internal resource needs, and achieve safe and stable operation of the virtual power plant. Second, to collect information on lower-level user-side resources through intelligent terminal devices, and flexibly coordinate lower-level resources based on actual conditions and various external trends, formulating reasonable scheduling plans so that the virtual power plant operator and lower-level user-side resources can play their respective roles while also achieving their respective benefits. The functional positioning of the upper-level virtual power plant operator is as follows:

[0094] ① Attract multiple types of resources for aggregation and collaboration, and provide support and platforms for them. On the one hand, provide access platforms for widely distributed renewable energy generators and build a standardized access system; on the other hand,

[0095] Virtual power plant operators can act as agents for users with smaller energy demands, allowing them to also enjoy preferential electricity prices. For users with larger energy demands, operators focus more on tracking and regulating their electricity usage behavior, which can not only help large users reduce their electricity costs, but also develop their regulation potential.

[0096] ② As a leader, the virtual power plant operator controls and manages lower-level user-side resources to ensure a balanced supply and demand of electricity. Relying on a cloud-edge-end collaborative framework, the virtual power plant operator receives and analyzes relevant operational data from lower-level user-side resources, optimizing operations according to set goals. When energy supply is insufficient, the operator can direct lower-level adjustable resources to respond or purchase electricity from the external main grid via interconnection lines, but the maximum allowable value of the interconnection lines cannot be exceeded. Furthermore, the virtual power plant operator needs to continuously monitor equipment status and energy flow. In the event of emergencies or extreme weather, the operator can promptly adjust internal resources to ensure a balanced supply and demand and mitigate risks.

[0097] ③ Coordinate internal resources to improve overall benefits. On the one hand, virtual power plant operators, taking into account their own needs and the external environment, decide what resources to aggregate, the quantity of aggregated resources, and the ratio between various resources to ensure the safe and reliable operation of the virtual power plant, while also avoiding resource waste and reducing costs. On the other hand, during the optimized operation process, by improving the accuracy of distributed renewable energy output forecasts, reducing the impact of the uncertainty of distributed renewable energy output, and by rationally coordinating other adjustable resources in the lower layer to improve the overall regulation capacity of the virtual power plant, the safe and stable operation of the virtual power plant and the improvement of overall benefits are guaranteed.

[0098] 2. At the lower level, the main resource components are user-side resources, which exist in various forms such as consumers and prosumers. Generally, driven by reasons such as small capacity, obstructed access to the grid, and high energy costs, they participate in the aggregation of virtual power plants and accept the regulation of upper-level operators. They respond according to their actual conditions, play their own role, and improve overall benefits. For supply resources, the goal is to hope that operators can reasonably arrange power generation plans and recover power generation costs; for load resources, the goal is to obtain a stable, reliable, and economical power supply and reduce electricity purchase costs. The functional positioning of each entity at the lower level is as follows:

[0099] ① Distributed power generation resources: The main function of distributed power generation resources is to supply electrical energy to the internal loads of the virtual power plant. In addition, some controllable units, such as gas-fired units, also serve as adjustable resources. When fluctuations occur, the virtual power plant operator can regulate them by issuing instructions while meeting other requirements such as ramp constraints.

[0100] ② Energy storage resources: Energy storage resources typically act as a buffer between the source and the load. When the overall supply of a virtual power plant exceeds demand, the upper-level virtual power plant operator regulates the energy storage resources to charge, and energy storage then acts as a load within the virtual power plant. When demand exceeds supply, the operator regulates the energy storage resources to discharge, and energy storage then acts as a power source within the virtual power plant. However, during this regulation process, it is important to note the constraints of energy storage capacity and the requirement that the energy storage charge and discharge power must be equal throughout the day.

[0101] ③ Load resources (including third-party load integrators and large users): As the most regulated resources among the lower-level resources, load resources primarily respond to incentive signals from the upper layer and adjust their electricity usage based on their own circumstances. On the one hand, by participating in demand response, load resources can reduce their overall electricity costs. On the other hand, the regulation capacity of load resources largely determines the overall flexibility of the virtual power plant, helping to improve the overall economic and clean benefits of the virtual power plant.

[0102] Based on the aforementioned cloud-edge-device collaboration framework, the mechanism for regulating and controlling virtual power plants can be illustrated as follows:

[0103] The basic logic of the virtual power plant control mechanism based on cloud-edge-end collaboration is that the cloud center layer receives data and information reported, processed, and analyzed layer by layer starting from the terminal layer through the edge layer, and determines the optimization goals based on the virtual power plant's analysis of various internal and external influencing factors. The scheduling plan is determined based on the stored calculation logic, and the task is sent to the edge layer, which is then decomposed from the edge layer to the terminal layer.

[0104] ① Edge-end collaboration: After collecting real-time data and information on the status of distributed generators, energy storage equipment, and load resources, intelligent terminal devices deployed at the terminal layer upload this data and information, including the environmental conditions of the distributed resources, the operating status of each device, and user electricity usage, to the edge layer. Based on the real-time data, information, and historical data, the edge layer performs preliminary calculations and analysis using its built-in computing modules. Based on actual needs, the edge layer either directly feeds the results back to the terminal layer or waits for further calculations and analysis from the cloud center layer before passing them to the terminal layer through analysis by the edge layer.

[0105] ② Edge-Cloud Collaboration: After processing terminal-layer data and information, the edge layer uploads the required data and information to the cloud center layer, storing the remaining information according to rules for immediate access. Based on the received data and information, the cloud center layer uses internally stored calculation rules to perform predictions and optimization based on the virtual power plant's goals or needs, and finally converts the results into instructions and sends them to the edge layer.

[0106] ③ Cloud collaboration: For some important situations or emergencies, the cloud center layer needs to directly regulate the terminal layer. At this time, all calculations and transmissions do not pass through the edge layer. The data and information collected by the terminal layer are directly transmitted to the cloud center layer. At the same time, the cloud center layer directly issues instructions to the terminal layer, thereby responding to the situation of the virtual power plant handling emergencies.

[0107] The predicted output of the above-mentioned wind turbines can be predicted using a neural network.

[0108] CNNs feature local connections and weight sharing between layers, and their convolution and pooling layers effectively extract data features and reduce their dimensionality, resulting in more stable data characteristics. However, these characteristics lead CNNs to prioritize capturing the spatial morphology and image features of data vectors, potentially overlooking the continuity and evolution of distributed wind turbine data in the temporal dimension during learning. However, the strong memory of LSTMs gives them a significant advantage in learning long time series problems.

[0109] Based on the above two considerations, the present invention proposes to construct a distributed wind turbine output power prediction model based on a CNN-LSTM fusion neural network, in which CNN is mainly used to capture the morphological characteristics of the distributed wind turbine data curve, and LSTM is used to capture the temporal characteristics of the distributed wind turbine data curve.

[0110] like Figure 3 As shown, the input of the combined prediction model constructed by the present invention is the characteristic data, and the output is the hourly wind turbine output power on the target day:

[0111] The first step is to input feature data into the input layer, primarily weather and time information. Weather information includes wind speed, wind direction, and air density; time information uses the month and day as feature data. This input is represented as a two-dimensional data matrix, with the weather and time feature data displayed horizontally and the time series displayed vertically.

[0112] The second step is the convolutional layer, the core component of the CNN. Its primary function is to fully extract data from the input layer. Multiple convolution kernels are used in this layer to extract data features, effectively capturing the nonlinear relationship between factors influencing output and wind turbine output.

[0113] The third step is the pooling layer. At this point, the extracted data may still be of high dimensionality. This requires filtering and dimensionality reduction through the pooling layer to remove noise and uncertainty. Alternatively, the data can be processed through multiple convolutional and pooling layers.

[0114] In the fourth step, the processed data is input into the LSTM through the fully connected layer for time series learning. To analyze and evaluate the performance of the model, this article uses the test set data as a sample to test and analyze the model performance after completing the model training, and uses the following three indicators to quantify the model prediction performance:

[0115] ① Mean absolute percentage error (MAPE)

[0116]

[0117] ②Root mean square error (RMSE)

[0118]

[0119] ③ Mean absolute error (MSE)

[0120]

[0121] in is the actual value; y i is the predicted value; n is the total number of samples.

[0122] Figure 4 This is a structural block diagram of a multi-time-scale scheduling device based on a virtual power plant according to an embodiment of the present invention.

[0123] like Figure 4 As shown, the multi-time scale scheduling device based on the virtual power plant includes:

[0124] A first function construction module 410 is configured to construct a day-ahead objective function with the goal of maximizing the revenue of the virtual power plant's day-ahead power dispatch plan for the next day, and with the condition that each power device in the virtual power plant responds to the first type of load and does not respond to the second type of load;

[0125] A first optimization module 420 is configured to maximize the day-ahead objective function based on the day-ahead dispatch constraints of the virtual power plant, thereby obtaining a day-ahead dispatch curve for each of the power equipment and the first type of load on the next day, wherein the power equipment includes a gas generator set, a wind turbine set, a photovoltaic generator set, and an energy storage device;

[0126] A first scheduling module 430 is configured to perform power scheduling for the virtual power plant on the next day based on the day-ahead scheduling curve, and monitor the actual output of the wind turbine and photovoltaic generator sets in various time periods on the next day;

[0127] A second function construction module 440 is configured to construct an intraday objective function based on the objective of minimizing the adjustment cost of the intraday power dispatch plan of the virtual power plant in the first time period, and conditional on the gas-fired units in the virtual power plant responding to the second type of load, when the actual output of the wind turbine and / or photovoltaic unit in the first time period of the next day does not match the predicted output for the corresponding time period in the corresponding day-ahead dispatch curve;

[0128] The second optimization module 450 is used to solve the minimum value of the intraday objective function based on the intraday scheduling constraints of the virtual power plant, and obtain the intraday scheduling curve of the gas unit and the first type of load in the first time period.

[0129] In one embodiment, the first function building module includes:

[0130] a revenue function determining unit, configured to determine a revenue function for the virtual power plant to sell electricity to internal loads on the next day based on the upward power regulation and downward power regulation of the first type of load at various times on the next day and the electricity sales price of the virtual power plant at various times on the next day;

[0131] A day-ahead objective function determination unit is configured to determine the day-ahead objective function based on the revenue function, the gas-fired power generation cost, wind power generation cost, photovoltaic power generation cost, solar curtailment penalty cost, wind curtailment penalty cost, and energy storage charging and discharging cost of the virtual power plant, as well as the electric energy interaction cost between the virtual power plant and the power grid.

[0132] In one embodiment, the first type of load is a load that is notified one day in advance of participating in the regulation, and the electricity price of the first type of load is determined based on a price coefficient corresponding to the load change rate;

[0133] The second type of load is a load that is notified of participating in regulation in advance for a period of time, and the electricity price of the second type of load is an electricity price determined based on the load adjustment power of the second type of load.

[0134] In one embodiment, the profit function is:

[0135]

[0136] in, represents the total revenue of the virtual power plant from selling electricity to internal loads on the next day, represents the upward adjustment power of the first type of load at time t, represents the downward adjustment power of the first type of load at time t, D t represents the unregulated power of all loads in the virtual power plant, p l,t represents the time-of-use electricity price of the virtual power plant supplying power to the internal load at time t, Δp l,t It represents the price fluctuation of the time-of-use electricity price of the virtual power plant supplying power to the internal load at time t.

[0137] In one embodiment, the calculation formula for the unadjusted power of all loads in the virtual power plant is as follows:

[0138]

[0139] in, represents the unadjusted power of the first type of load at time t, represents the unadjusted power of the second type of load at time t, represents the rigid load in the virtual power plant.

[0140] In one embodiment, the intraday objective function is:

[0141]

[0142] Where C represents the intraday objective function, ΔC gasrepresents the cost of adjusting the gas-fired power generation output of the gas-fired unit, ΔC grid represents the cost of the virtual power plant temporarily purchasing electricity from the grid, represents the adjustment cost of the second type of load, ΔP gas,t represents the adjusted power of the gas turbine at time t, Δp gas,t represents the marginal cost of output adjustment of the gas turbine at time t, represents the power purchased by the virtual power plant from the grid at time t; represents the price of electricity purchased by the virtual power plant from the grid at time t, Indicates the adjusted power of the second type of load, represents the subsidy cost of the second type of load, and Δt2 represents the duration of the first time period.

[0143] For the description of specific functions and examples of each module and submodule of the system in the embodiment of the present invention, please refer to the relevant description of the corresponding steps in the above method embodiment, which will not be repeated here.

[0144] In the technical solution of the present invention, the acquisition, storage and application of user personal information involved are in compliance with the provisions of relevant laws and regulations and do not violate public order and good morals.

[0145] According to an embodiment of the present invention, the present invention further provides a system and a readable storage medium.

[0146] Figure 5 A schematic block diagram of an example electronic device 800 that can be used to implement embodiments of the present invention is shown. The electronic device is intended to represent various forms of digital computers, such as laptop computers, desktop computers, workstations, personal digital assistants, servers, blade servers, mainframe computers, and other suitable computers. The electronic device can also represent various forms of mobile devices, such as personal digital assistants, cellular phones, smartphones, wearable devices, and other similar computing devices. The components shown herein, their connections and relationships, and their functions are merely examples and are not intended to limit the implementation of the present invention described and / or claimed herein.

[0147] like Figure 5 As shown, the device 800 includes a computing unit 801, which can perform various appropriate actions and processes according to a computer program stored in a read-only memory (ROM) 802 or a computer program loaded from a storage unit 808 into a random access memory (RAM) 803. Various programs and data required for the operation of the device 800 can also be stored in the RAM 803. The computing unit 801, the ROM 802, and the RAM 803 are connected to each other via a bus 804. An input / output (I / O) interface 805 is also connected to the bus 804.

[0148] Various components in device 800 are connected to I / O interface 805, including an input unit 806, such as a keyboard, mouse, etc.; an output unit 807, such as various types of displays, speakers, etc.; a storage unit 808, such as a magnetic disk, optical disk, etc.; and a communication unit 809, such as a network card, modem, wireless communication transceiver, etc. The communication unit 809 allows device 800 to exchange information / data with other devices via a computer network such as the Internet and / or various telecommunication networks.

[0149] The computing unit 801 can be any general-purpose and / or specialized processing component with processing and computing capabilities. Some examples of the computing unit 801 include, but are not limited to, a central processing unit (CPU), a graphics processing unit (GPU), various specialized artificial intelligence (AI) computing chips, various computing units that run machine learning model algorithms, a digital signal processor (DSP), and any appropriate processor, controller, microcontroller, etc. The computing unit 801 performs the various methods and processes described above, such as the multi-timescale scheduling method based on a virtual power plant. For example, in some embodiments, the multi-timescale scheduling method based on a virtual power plant can be implemented as a computer software program that is tangibly contained in a machine-readable medium, such as a storage unit 808. In some embodiments, part or all of the computer program can be loaded and / or installed on the device 800 via the ROM 802 and / or the communication unit 809. When the computer program is loaded into the RAM 803 and executed by the computing unit 801, one or more steps of the multi-timescale scheduling method based on a virtual power plant described above can be performed. Alternatively, in other embodiments, the computing unit 801 may be configured in any other appropriate manner (for example, by means of firmware) to execute a multi-time-scale scheduling method based on a virtual power plant.

[0150] Various embodiments of the systems and techniques described herein can be implemented in digital electronic circuit systems, integrated circuit systems, field programmable gate arrays (FPGAs), application specific integrated circuits (ASICs), application specific standard products (ASSPs), system-on-chip systems (SOCs), programmable logic devices (CPLDs), computer hardware, firmware, software, and / or combinations thereof. These various embodiments can include being implemented in one or more computer programs that are executable and / or interpreted on a programmable system comprising at least one programmable processor, which can be a special purpose or general purpose programmable processor that can receive data and instructions from a storage system, at least one input device, and at least one output device, and transmit data and instructions to the storage system, the at least one input device, and the at least one output device.

[0151] The program code for implementing the method of the present invention can be written in any combination of one or more programming languages. Such program code can be provided to a processor or controller of a general-purpose computer, a special-purpose computer, or other programmable data processing device so that when the program code is executed by the processor or controller, the functions / operations specified in the flow chart and / or block diagram are implemented. The program code can be executed entirely on the machine, partially on the machine, as a stand-alone software package, partially on the machine and partially on a remote machine, or entirely on a remote machine or server.

[0152] In the context of the present invention, machine-readable medium can be a tangible medium that can contain or store a program for use with an instruction execution system, device or equipment or used in combination with an instruction execution system, device or equipment. Machine-readable medium can be a machine-readable signal medium or a machine-readable storage medium. Machine-readable medium can include, but is not limited to, electronic, magnetic, optical, electromagnetic, infrared or semiconductor systems, devices or equipment, or any suitable combination of the foregoing. More specific examples of machine-readable storage media can include electrical connections based on one or more lines, portable computer disks, hard disks, random access memories (RAM), read-only memories (ROM), erasable programmable read-only memories (EPROM or flash memory), optical fibers, portable compact disk read-only memories (CD-ROM), optical storage devices, magnetic storage devices, or any suitable combination of the foregoing.

[0153] To provide interaction with a user, the systems and techniques described herein can be implemented on a computer having: a display device (e.g., a CRT (cathode ray tube) or LCD (liquid crystal display) monitor) for displaying information to the user; and a keyboard and pointing device (e.g., a mouse or trackball) through which the user can provide input to the computer. Other types of devices can also be used to provide interaction with the user; for example, the feedback provided to the user can be any form of sensory feedback (e.g., visual feedback, auditory feedback, or tactile feedback); and input from the user can be received in any form (including acoustic input, voice input, or tactile input).

[0154] The systems and techniques described herein can be implemented in a computing system that includes back-end components (e.g., as a data server), or a computing system that includes middleware components (e.g., an application server), or a computing system that includes front-end components (e.g., a user computer having a graphical user interface or a web browser through which a user can interact with implementations of the systems and techniques described herein), or a computing system that includes any combination of such back-end components, middleware components, or front-end components. The components of the system can be interconnected by any form or medium of digital data communication (e.g., a communication network). Examples of communication networks include a local area network (LAN), a wide area network (WAN), and the Internet.

[0155] A computer system may include a client and a server. The client and server are generally remote from each other and typically interact through a communication network. The client-server relationship arises through computer programs running on the respective computers and having a client-server relationship with each other. The server may be a cloud server, a server in a distributed system, or a server integrated with a blockchain.

[0156] It should be understood that the various forms of the processes shown above can be used to reorder, add, or delete steps. For example, the steps described in the present invention can be performed in parallel, sequentially, or in a different order, as long as the desired results of the technical solutions disclosed in the present invention can be achieved. This is not limited herein.

[0157] The above specific embodiments do not limit the scope of protection of the present invention. Those skilled in the art will appreciate that various modifications, combinations, sub-combinations, and substitutions may be made based on design requirements and other factors. Any modifications, equivalent substitutions, and improvements made within the principles of the present invention are intended to be included within the scope of protection of the present invention.

Claims

1. A multi-time scale scheduling method based on virtual power plant, characterized in that: include: A day-ahead objective function is constructed with the goal of maximizing the revenue of the virtual power plant's next-day day-ahead power dispatch plan and with the condition that each power device in the virtual power plant responds to the first type of load and does not respond to the second type of load; Based on the day-ahead dispatch constraints of the virtual power plant, solving the day-ahead objective function for a maximum value to obtain a day-ahead dispatch curve for each of the power equipment and the first type of load on the next day, wherein the power equipment includes a gas generator set, a wind generator set, a photovoltaic generator set, and an energy storage device; Based on the day-ahead dispatch curve, the virtual power plant is dispatched for power on the next day, and the actual output of the wind turbine and photovoltaic unit is monitored in each time period on the next day; In a case where the actual output of the wind turbine generator set and / or the photovoltaic generator set in the first time period of the next day does not match the predicted output for the corresponding time period in the corresponding day-ahead dispatch curve, an intraday objective function is constructed with the goal of minimizing the adjustment cost of the intraday power dispatch plan of the virtual power plant in the first time period and with the condition that the gas generator set in the virtual power plant responds to the second type of load; Based on the intraday scheduling constraints of the virtual power plant, the minimum value of the intraday objective function is solved to obtain the intraday scheduling curve of the gas unit and the first type of load in the first time period.

2. The method according to claim 1, characterized in that The first type of load is a load that is notified one day in advance to participate in the regulation, and the electricity price of the first type of load is determined based on the price coefficient corresponding to the load change rate; The second type of load is a load that is notified of participating in regulation in advance for a period of time, and the electricity price of the second type of load is an electricity price determined based on the load adjustment power of the second type of load.

3. The method according to claim 1, characterized in that The goal is to maximize the profit of the virtual power plant's day-ahead power dispatch plan on the next day, and the condition is that each power device in the virtual power plant responds to the first type of load and does not respond to the second type of load. The day-ahead objective function is constructed, including: Determine a revenue function for the virtual power plant to sell electricity to internal loads on the next day based on the upward power regulation and downward power regulation of the first type of load at various times on the next day and the electricity sales price of the virtual power plant at various times on the next day; The day-ahead objective function is determined based on the revenue function, the gas-fired power generation cost, wind power generation cost, photovoltaic power generation cost, solar curtailment penalty cost, wind curtailment penalty cost and energy storage charging and discharging cost of the virtual power plant, as well as the electric energy interaction cost between the virtual power plant and the power grid.

4. The method according to claim 3, characterized in that The profit function is: in, represents the total revenue of the virtual power plant from selling electricity to internal loads on the next day, represents the upward adjustment power of the first type of load at time t, represents the downward adjustment power of the first type of load at time t, D t represents the unregulated power of all loads in the virtual power plant, p l,t represents the time-of-use electricity price of the virtual power plant supplying power to the internal load at time t, Δp l,t It represents the price fluctuation of the time-of-use electricity price of the virtual power plant supplying power to the internal load at time t.

5. The method according to claim 4, characterized in that The calculation formula for the unadjusted power of all loads in the virtual power plant is as follows: in, represents the unadjusted power of the first type of load at time t, represents the unadjusted power of the second type of load at time t, represents the rigid load in the virtual power plant.

6. The method according to claim 1, characterized in that The intraday objective function is: Where C represents the intraday objective function, ΔC gas represents the cost of adjusting the gas-fired power generation output of the gas-fired unit, ΔC grid represents the cost of the virtual power plant temporarily purchasing electricity from the grid, represents the adjustment cost of the second type of load, ΔP gas,t represents the adjusted power of the gas turbine at time t, Δp gas,t represents the marginal cost of output adjustment of the gas turbine at time t, represents the power purchased by the virtual power plant from the grid at time t; represents the price of electricity purchased by the virtual power plant from the grid at time t, Indicates the adjusted power of the second type of load, represents the subsidy cost of the second type of load, and Δt2 represents the duration of the first time period.

7. A multi-time scale scheduling device based on a virtual power plant, characterized in that: include: The first function construction module is configured to construct a day-ahead objective function with the goal of maximizing the revenue of the virtual power plant's day-ahead power dispatch plan on the next day and with the condition that each power device in the virtual power plant responds to the first type of load and does not respond to the second type of load; a first optimization module, configured to solve the day-ahead objective function for a maximum value based on the day-ahead dispatch constraints of the virtual power plant, and obtain a day-ahead dispatch curve for each of the power equipment and the first type of load on the next day, wherein the power equipment includes a gas generator set, a wind turbine set, a photovoltaic generator set, and an energy storage device; a first scheduling module, configured to perform power scheduling for the virtual power plant on the next day based on a day-ahead scheduling curve, and monitor actual outputs of the wind turbine generator set and the photovoltaic generator set in various time periods on the next day; a second function construction module for constructing an intraday objective function with the goal of minimizing the adjustment cost of the intraday power dispatch plan of the virtual power plant in the first time period, and with the gas-fired units in the virtual power plant responding to the second type of load as a condition, when the actual output results of the wind turbine generator set and / or the photovoltaic generator set in the first time period of the next day do not match the predicted output for the corresponding time period in the corresponding day-ahead dispatch curve; The second optimization module is used to solve the minimum value of the intraday objective function based on the intraday scheduling constraints of the virtual power plant, and obtain the intraday scheduling curve of the gas unit and the first type of load in the first time period.

8. The device according to claim 1, characterized in that The first function building module includes: a revenue function determining unit, configured to determine a revenue function for the virtual power plant to sell electricity to internal loads on the next day based on the upward power regulation and downward power regulation of the first type of load at various times on the next day and the electricity sales price of the virtual power plant at various times on the next day; A day-ahead objective function determination unit is configured to determine the day-ahead objective function based on the revenue function, the gas-fired power generation cost, wind power generation cost, photovoltaic power generation cost, solar curtailment penalty cost, wind curtailment penalty cost, and energy storage charging and discharging cost of the virtual power plant, as well as the electric energy interaction cost between the virtual power plant and the power grid.

9. A multi-timescale scheduling system based on a virtual power plant, comprising: at least one processor; as well as a memory communicatively connected to the at least one processor; wherein, The memory stores instructions that can be executed by the at least one processor, and the instructions are executed by the at least one processor to enable the at least one processor to perform the method according to any one of claims 1 to 6.

10. A non-transitory computer-readable storage medium storing computer instructions, wherein: The computer instructions are used to cause the computer to execute the method according to any one of claims 1 to 6.

Images