Multi-virtual power plant intelligent scheduling decision-making method and system considering uncertain risk

Abstract

The invention discloses a multi-virtual power plant intelligent scheduling decision-making method and system considering uncertain risks, and the method comprises the steps: constructing an objective function, and comprehensively optimizing the response degree, the total response cost and the uncertain risks of provincial scheduling demands; setting constraint conditions in the virtual power plant intelligent scheduling decision model, wherein the constraint conditions comprise setting of upper and lower limit constraints of response power, setting of response rate constraints, setting of response time constraints and setting of load state constraints of an energy storage system; data collection and preprocessing: the virtual power plant collects diversified data through a user layer measurement unit, and a decision-making layer preprocesses the data; and scheduling and adjusting in real time. The model proposed by the invention not only ensures that each virtual power plant can maximize own profit when participating in market bidding through an upper layer optimization strategy quotation coefficient, but also gives consideration to the stability, safety and economic benefits of the whole power grid, and the design ingeniously balances the local benefit of a single VPP and the global optimality of the power grid, and improves the market bidding efficiency. And the decision-making process is more scientific and reasonable.

Description

Technical Field

[0001] The present invention relates to a method and system for intelligent dispatching decision-making of a virtual power plant, and in particular to a method and system for intelligent dispatching decision-making of multiple virtual power plants taking into account uncertain risks. Background Art

[0002] With the transformation of the global energy structure and the opening of the electricity market, the importance of virtual power plant technology as a new benchmark for distributed energy management is becoming increasingly prominent. Through advanced information and communication technologies and software systems, virtual power plants efficiently integrate distributed power sources, energy storage systems, controllable loads and other resources to form a flexible and controllable virtual power supply complex, deeply participating in the electricity market and ancillary service market. This technology not only optimizes resource allocation and reduces system operating costs, but also significantly improves the flexibility and resilience of the energy system, ensuring the stability and security of power supply in a changing market demand and uncertain energy supply environment. Therefore, virtual power plant technology is an important force in promoting the energy revolution, building a new power system, and achieving efficient energy utilization and sustainable development.

[0003] When exploring the efficient operation mode of virtual power plants, it is particularly important to analyze and build a reasonable intelligent scheduling decision model. This model is the core support for the efficient operation and flexible scheduling of virtual power plants. It achieves the optimal configuration and efficient utilization of resources through in-depth analysis and intelligent optimization of massive and dispersed distributed energy resources. The model can predict and respond to changes in grid demand, ensure the balance of power supply and demand, and improve the flexibility and resilience of the grid. In addition, the intelligent scheduling decision model also promotes the deep integration of virtual power plants and the power market, creates favorable conditions for virtual power plants to participate in market bidding and provide ancillary services, and further promotes the commercial application and sustainable development of virtual power plant technology.

[0004] However, under the current "layered and zoned" power dispatching management system framework in my country, the construction of a virtual power plant intelligent dispatching decision model faces multiple challenges. First, the division of responsibilities between dispatching centers at all levels for power grid security and reserve capacity management is clear, but how to efficiently configure the reserve demand capacity of the entire system and each zone and realize the effective dispatch of reserve between zones has become a key issue that needs to be solved urgently. This requires the model to have a high degree of coordination and flexibility to adapt to the complex interaction and dynamic adjustment between dispatching centers at different levels. Secondly, most existing studies are limited to treating VPP (virtual power plant) as a price acceptor, and only passively respond to the decision-making of the upper power grid to adjust the internal operating state, ignoring the potential of VPP as an independent market competition subject under market conditions. With the continuous maturity and opening of the power market, VPP is eager to actively participate in the power grid ancillary service market and strive for higher economic benefits with active strategies. Therefore, how to fully integrate the market mechanism into the model and consider the autonomous decision-making ability and market competition strategy of VPP has become another major challenge that must be faced when building an intelligent dispatching decision model. Summary of the invention

[0005] Purpose of the invention: The purpose of the invention is to provide a multi-virtual power plant intelligent scheduling decision-making method and system taking into account uncertain risks to achieve optimal allocation of power resources.

[0006] Technical solution: The present invention includes the following steps: constructing an objective function to comprehensively optimize the responsiveness, total response cost and uncertainty risk of provincial dispatching demand; setting constraints in the intelligent dispatching decision model of the virtual power plant, including setting upper and lower limit constraints of the response power, setting response rate constraints, setting response time constraints, and setting load state constraints of the energy storage system; data collection and preprocessing, the virtual power plant collects diversified data through user-level measurement units, and the decision-making layer preprocesses data; real-time scheduling and adjustment.

[0007] The objective function is expressed as:

[0008] minf=ω1f response +ω2f cost +ω3f risk

[0009] In the formula, ω1, ω2, ω3 represent weight coefficients, f response 、f cost 、f risk They represent the objective functions of provincial dispatch demand satisfaction, dispatch cost and dispatch risk respectively.

[0010] The total response cost means that when there are multiple resources in the power dispatching area that can respond to the provincial dispatching, the dispatching units with lower dispatching costs should be given priority to ensure that the total dispatching response cost is minimized, that is, minf cost ,

[0011]

[0012] In the formula, t start is the scheduling start time, t end The scheduling end time; Represents the unit cost of scheduling response of the i-th resource.

[0013] The uncertainty risk refers to the screening of resources according to the historical response of dispatchable resources in the dispatch decision process, giving priority to dispatching resources with high response stability, introducing dispatch risk cost, and ensuring the minimum risk cost, that is, minf risk ,

[0014]

[0015] In the formula, C risk represents the unit energy scheduling risk cost, Represents the response uncertainty of the i-th scheduling resource, which is 0 initially.

[0016] The upper and lower limits of the response power are set to define the maximum power and minimum power range that the VPP can provide:

[0017]

[0018] Where p i (t) Represents the corresponding power, adjustable lower limit and upper limit of the i-th scheduling unit at time t.

[0019] The response time constraint is set as:

[0020]

[0021] In the formula, t start Indicates the time when the scheduling demand starts, t end Indicates the time when the scheduling demand ends, and t0 indicates the time when the scheduling demand instruction is issued; Indicates predicted output data.

[0022] The load state constraint of the energy storage system is set as:

[0023] 0≤SOC i (t)*V i -p i (t)≤V i

[0024] When the i-th dispatch resource is energy storage, the energy storage state of charge constraint should also be met. i (t) represents the state of charge of the energy storage at time t, V iRepresents the installed capacity of energy storage.

[0025] The real-time dispatch and adjustment are specifically as follows: in the virtual power plant intelligent dispatch system, the regional ancillary service market and the electric energy market are integrated, and the day-ahead pre-clearing is based on the principles of demand and sequence, and is sorted according to the unit price of deep peak regulation for intra-day use.

[0026] In the real-time scheduling and adjustment, the system dynamically adjusts the output of each unit according to the energy output demand and the market transaction settlement process to achieve optimal allocation of resources.

[0027] An intelligent dispatching decision system for multiple virtual power plants considering uncertain risks, comprising:

[0028] Objective function building module: When building an intelligent dispatching decision model for the collaboration between the virtual power plant and the regional power grid dispatching center, its core objective function aims to comprehensively optimize: the degree of response to provincial dispatching demand, the total response cost, and the uncertainty risk;

[0029] Constraint setting module in the intelligent dispatch decision model of virtual power plants: When building the intelligent dispatch decision model of virtual power plants, a series of constraints need to be strictly set;

[0030] Data collection and preprocessing module: The virtual power plant collects diverse data through user-level measurement units as the basis for sensing, predicting and controlling distributed generation and controllable resources;

[0031] Real-time dispatching and adjustment module: In the intelligent dispatching system of the virtual power plant, the regional ancillary service market is integrated with the electric energy market, and the day-ahead pre-clearing is based on the principle of demand and sequence, and is sorted according to the unit price of deep peak regulation for intraday use.

[0032] Beneficial effects: The present invention has the following advantages:

[0033] (1) The model proposed in this paper not only ensures that each virtual power plant can maximize its own profit when participating in the market bidding, but also takes into account the stability, security and economic benefits of the entire power grid through the upper-level optimization strategy quotation coefficient. This design cleverly balances the local interests of a single VPP with the optimality of the entire power grid, making the decision-making process more scientific and reasonable. At the same time, the multi-objective characteristics of the model further improve the economy and operation efficiency of the power system, providing strong support for the sustainable development of the power market.

[0034] (2) The multi-objective multi-virtual power plant collaborative decision-making system constructed by the present invention significantly enhances the flexibility and responsiveness of the power system. The system can comprehensively consider multiple dimensions such as provincial dispatching demand response and VPP uncertainty risk to achieve optimal allocation of power resources. The verification of engineering practice has proved the practicality and economy of the decision-making strategy in actual application, providing reliable guarantee for the stable operation of the power system and the healthy development of the power market. BRIEF DESCRIPTION OF THE DRAWINGS

[0035] Figure 1 It is a flowchart of the intelligent decision-making process of a virtual power plant according to an embodiment of the present invention;

[0036] Figure 2 It is a schematic diagram of realizing the intelligent decision-making project of a virtual power plant according to an embodiment of the present invention. DETAILED DESCRIPTION

[0037] The present invention will be further described below in conjunction with the accompanying drawings.

[0038] The multi-virtual power plant intelligent dispatching decision method considering uncertain risks of the present invention comprises the following steps:

[0039] Step 1, construct the objective function. When constructing the intelligent dispatch decision model of the virtual power plant (VPP) and the regional power grid dispatch center, its core objective function aims to comprehensively optimize the three key factors of "the degree of response to provincial dispatch demand", "the total cost of response" and "the risk of uncertainty" to respond to the power grid dispatch demand efficiently, economically and robustly. Among them, the objective function is expressed as:

[0040] minf=ω1f response +ω2f cost +ω3f risk

[0041] Where ω1, ω2, ω3 represent weight coefficients, f response 、f cost 、f risk They represent the objective functions of provincial dispatch demand satisfaction, dispatch cost and dispatch risk respectively.

[0042] The provincial dispatch demand satisfaction in the objective function is the dispatch demand that is first considered in the construction of the virtual power plant intelligent decision-making dispatch optimization model. If the dispatch cycle is 15 minutes, the sum of the response levels of each cycle segment is the largest in all durations. When constructing the objective function, you can consider setting a penalty function. If the conditions are not met, a penalty is imposed to ensure that the overall penalty value is minimized, that is, minf response .

[0043]

[0044] In the formula, C punish(t) represents the penalty unit price for failure to meet the provincial dispatch demand response in time period t (in order to ensure the dispatch demand, units with fast response time are encouraged to respond as soon as possible, and the earlier the time, the greater the penalty amount); P demand (t) represents the dispatch demand (gap value) of the provincial dispatch in time period t; p i (t) is the optimization decision variable, i.e., the power of the dispatch unit i in time period t (positive for output, negative for storage or load); It represents the predicted power of scheduling unit i in time period t.

[0045] The total response cost means that when there are multiple resources in the power dispatching area that can respond to the provincial dispatching, the dispatching unit with lower dispatching cost should be given priority to ensure that the total dispatching response cost is minimized, that is, minf cost .

[0046]

[0047] Where t start is the scheduling start time, t end The scheduling end time; Represents the unit cost of scheduling response of the i-th resource.

[0048] Uncertainty risk refers to the fact that in the actual dispatching process, the dispatching resources may not be executed according to the strategy, or there may be deviations from the strategy execution, which will cause the provincial dispatching to have to take some temporary measures to call for thermal power and other balances or directly lead to instability of the power grid. Therefore, in the dispatching decision-making process, resources should be screened according to the historical response of the dispatchable resources, and priority should be given to dispatching the control resources with high response stability. Introduce the dispatching risk cost to ensure the minimum risk cost, that is, minf risk .

[0049]

[0050] In the formula, C risk represents the unit energy scheduling risk cost, Represents the response uncertainty of the i-th scheduling resource, which is 0 initially.

[0051] Step 2: Set the constraints in the intelligent dispatch decision model of the virtual power plant. When building the intelligent dispatch decision model of the virtual power plant, a series of constraints must be strictly set to ensure its safe and efficient operation. Including:

[0052] (1) Set upper and lower limits of the response power to define the maximum and minimum power ranges that the VPP can provide; upper and lower limits of the response power refer to setting minimum and maximum limits on the power levels that a system or device can generate or respond to. This constraint is usually to ensure the stability, safety and efficiency of system operation. Specifically, the minimum power limit (lower limit) is used to prevent the system from being unable to work normally or reach the predetermined performance indicators due to too low power; while the maximum power limit (upper limit) is to avoid system overload, which may cause equipment damage, reduced energy efficiency or safety hazards.

[0053]

[0054] Where p i (t) Represents the corresponding power, adjustable lower limit and upper limit of the i-th scheduling unit at time t.

[0055] (2) Set a response rate constraint (i.e., ramp rate constraint) to limit the power adjustment speed of the VPP during the dispatch process and prevent excessive changes from causing impact on the power grid; the response rate constraint, also often referred to as the ramp rate constraint, is an important concept in the power system. This constraint mainly refers to the fact that the increase or decrease in the output power of the generator set in adjacent time periods must be kept within a certain limit range.

[0056]

[0057] (3) Set response time constraints to ensure that the VPP can quickly respond to grid dispatch instructions within the specified time; response time constraints mean that after the system or device receives an external request or instruction, it must complete the corresponding response action within the specified time range. This constraint ensures the timeliness and efficiency of the system or device, and avoids performance degradation or poor user experience caused by response delays.

[0058]

[0059] In the formula, t start Indicates the time when the scheduling demand starts, t end Indicates the time when the scheduling demand ends, and t0 indicates the time when the scheduling demand instruction is issued; Indicates predicted output data.

[0060] (4) Set the load state constraints of the energy storage system, monitor the charging and discharging state of the energy storage equipment, prevent overcharging or over-discharging, and ensure the long-term healthy operation of the energy storage equipment. The energy storage state of charge constraint is used to limit the percentage range of stored energy in the operation of the energy storage to ensure the stable operation of the energy storage system and extend its service life. Specifically, this constraint sets the lower and upper limits of the energy storage state of charge (SOC), that is, in actual operation of the energy storage system, the ratio of its stored power to the total energy storage capacity must be maintained within this range. Avoid too low SOC leading to deep discharge of energy storage, thereby shortening its service life; also avoid too high SOC leading to overcharging, which may cause damage to the energy storage equipment. Reasonable setting of the SOC constraint range can ensure the economical and reliable operation of the energy storage system.

[0061] 0≤SOC i (t)*V i -p i (t)≤V i

[0062] When the i-th dispatch resource is energy storage, the energy storage state of charge constraint should also be met. i (t) represents the state of charge of the energy storage at time t, V i Represents the installed capacity of energy storage.

[0063] Step 3: Data collection and preprocessing. The virtual power plant collects diverse data through user-layer measurement units as the basis for sensing, predicting, and controlling distributed power generation and controllable resources. Data is efficiently transmitted through the communication layer and protocol differences are processed to ensure accurate and timely information. The decision-making layer preprocesses data, eliminates anomalies, and improves data reliability, which is then used for regional flow calculations, load management, forecasting, and cost settlement, etc., to support intelligent scheduling decisions, optimize resource allocation, and maximize economic and social benefits.

[0064] Step 4: Real-time dispatch and adjustment. In the intelligent dispatch system of the virtual power plant, the regional ancillary service market is integrated with the electric energy market. The day-ahead pre-clearing is based on the principle of "on demand and in sequence" and is sorted according to the unit price of deep peak regulation (with response rate first) for intraday call. The units that must be started / shut down are adjusted based on the rolling combination and real-time margin, and the standby units are called. The on-site CFD settlement and the off-site unbalanced funds are channeled. Comprehensive energy demand and market transactions are used to generate economic dispatch decision plans.

[0065] The present invention also provides a multi-virtual power plant intelligent dispatching decision system considering uncertain risks, comprising:

[0066] Objective function construction module: When constructing an intelligent dispatching decision-making model that coordinates the virtual power plant (VPP) and the regional power grid dispatching center, its core objective function aims to comprehensively optimize the three key factors of "responsiveness to provincial dispatching demand", "total response cost" and "uncertainty risk" to respond to power grid dispatching needs efficiently, economically and robustly.

[0067] Constraint setting module in the intelligent dispatching decision model of virtual power plants: When building the intelligent dispatching decision model of virtual power plants, a series of constraints need to be strictly set to ensure its safe and efficient operation.

[0068] Data collection and preprocessing module: The virtual power plant collects diverse data through user-level measurement units as the basis for sensing, predicting and controlling distributed power generation and controllable resources. Data is efficiently transmitted through the communication layer and protocol differences are processed to ensure accurate and timely information.

[0069] Real-time dispatching and adjustment module: In the intelligent dispatching system of the virtual power plant, the regional ancillary service market is integrated with the electric energy market. The day-ahead pre-clearing is based on the principle of "on demand and in sequence" and is sorted according to the unit price of deep peak regulation for intraday use.

[0070] Example

[0071] This embodiment provides a multi-virtual power plant intelligent scheduling decision method considering uncertain risks, including the following steps:

[0072] Step 1: Construct the objective function. From the perspective of power grid management, considering certain constraints as an independent individual participating in power auxiliary services, and taking into account the impact of VPP's uncertain risks on the power grid, the objective function of VPP intelligent dispatching decision-making is constructed by considering three parts, namely "the degree of response to provincial dispatching demand", "the total response cost", and "the uncertainty risk".

[0073] minf=ω1f response +ω2f cost +ω3f risk

[0074] Where ω1, ω2, ω3 represent weight coefficients, f response 、f cost 、f risk They represent the objective functions of provincial dispatch demand satisfaction, dispatch cost and dispatch risk respectively.

[0075] The provincial dispatch demand satisfaction in the objective function is the dispatch demand that is first considered in the construction of the virtual power plant intelligent decision-making dispatch optimization model. If the dispatch cycle is 15 minutes, the sum of the response levels of each cycle segment is the largest in all durations. When constructing the objective function, you can consider setting a penalty function. If the conditions are not met, a penalty is imposed to ensure that the overall penalty value is minimized, that is, minf response .

[0076]

[0077] In the formula, C punish (t) represents the penalty unit price for failure to meet the provincial dispatch demand response in time period t (in order to ensure the dispatch demand, units with fast response time are encouraged to respond as soon as possible, and the earlier the time, the greater the penalty amount); P demand (t) represents the dispatch demand (gap value) of the provincial dispatch in time period t; p i (t) is the optimization decision variable, i.e., the power of the dispatch unit i in time period t (positive for output, negative for storage or load); It represents the predicted power of scheduling unit i in time period t.

[0078] The total response cost means that when there are multiple resources in the power dispatching area that can respond to the provincial dispatching, the dispatching unit with lower dispatching cost should be given priority to ensure that the total dispatching response cost is minimized, that is, minf cost .

[0079]

[0080] Where tstart is the scheduling start time, tend is the scheduling end time; Represents the unit cost of scheduling response of the i-th resource.

[0081] Uncertainty risk refers to the fact that in the actual dispatching process, the dispatching resources may not be executed according to the strategy, or there may be deviations from the strategy execution, which will cause the provincial dispatching to have to take some temporary measures to call for thermal power and other balances or directly lead to instability of the power grid. Therefore, in the dispatching decision-making process, resources should be screened according to the historical response of the dispatchable resources, and priority should be given to dispatching the control resources with high response stability. Introduce the dispatching risk cost to ensure the minimum risk cost, that is, minf risk .

[0082]

[0083] In the formula, C risk represents the unit energy scheduling risk cost, Represents the response uncertainty of the i-th scheduling resource, which is 0 initially.

[0084] Step 2: Set the constraints in the virtual power plant intelligent dispatch decision model

[0085] (1) Response power upper and lower limit constraints

[0086] The upper and lower limits of response power refer to setting the minimum and maximum limits on the power levels that a system or device can generate or respond to. This constraint is usually to ensure the stability, safety and efficiency of system operation. Specifically, the minimum power limit (lower limit) is used to prevent the system from being unable to work normally or reach the predetermined performance indicators due to too low power; while the maximum power limit (upper limit) is to avoid system overload, which may cause equipment damage, reduced energy efficiency or safety hazards.

[0087]

[0088] Where p i (t) Represents the corresponding power, adjustable lower limit and upper limit of the i-th scheduling unit at time t.

[0089] (2) Response rate constraint (climbing rate constraint)

[0090] Response rate constraint, also often called ramp rate constraint, is an important concept in power systems. This constraint mainly refers to the fact that the increase or decrease in the output power of the generator set in adjacent time periods must be kept within a certain limit.

[0091]

[0092] (3) Response time constraints

[0093] Response time constraints mean that after a system or device receives an external request or instruction, it must complete the corresponding response action within a specified time range. This constraint ensures the timeliness and efficiency of the system or device, and avoids performance degradation or poor user experience caused by response delays.

[0094]

[0095] In the formula, t start Indicates the time when the scheduling demand starts, t end Indicates the time when the scheduling demand ends, and t0 indicates the time when the scheduling demand instruction is issued; Indicates predicted output data.

[0096] (4) Energy storage load state constraints

[0097] The energy storage state of charge constraint is used to limit the percentage range of energy storage during operation to ensure the stable operation of the energy storage system and extend its service life. Specifically, this constraint sets the lower and upper limits of the energy storage state of charge (SOC), that is, in actual operation of the energy storage system, the ratio of its stored power to the total energy storage capacity must be maintained within this range. Avoid too low SOC leading to deep discharge of energy storage, thereby shortening its service life; also avoid too high SOC leading to overcharging, which may cause damage to the energy storage equipment. Reasonable setting of the SOC constraint range can ensure the economical and reliable operation of the energy storage system.

[0098] 0≤SOC i (t)*V i -p i (t)≤V i

[0099] When the i-th dispatch resource is energy storage, the energy storage state of charge constraint should also be met. i (t) represents the state of charge of the energy storage at time t, V i Represents the installed capacity of energy storage.

[0100] Step 3, data collection and input. Virtual power plants collect a wide range of data such as power generation curves, load changes, voltage and current, equipment status, and internal user power consumption plans through measurement units deployed at important nodes in the user layer. These data are the basis for virtual power plants to perceive, predict, and control the status of distributed power generation and controllable resources. Subsequently, these data are efficiently transmitted through the communication layer, and the differences in communication protocols between different devices are processed during the transmission process to ensure the accuracy and timeliness of the information. At the decision-making level, the virtual power plant system first pre-processes these massive data, including data cleaning, denoising, verification, and standardization to eliminate outliers and erroneous data and improve data reliability and availability. The pre-processed data is used for core functions such as regional power flow distribution calculation, controllable load management, power generation and load forecasting, and cost settlement, providing solid data support for the intelligent scheduling decision-making of virtual power plants.

[0101] The input parameters include two parts, namely, the demand information of the provincial dispatch and the dispatchable information of the virtual power plant.

[0102] The demand information of provincial dispatch is the dispatch demand information for a period of time in the future. demand (t), an example is as follows.

[0103] Table 1 Example of provincial dispatching demand

[0104]

[0105]

[0106] The dispatchable information of virtual power plants mainly includes: the type of dispatch unit; the forecast information of the adjustable capacity in the future ( That is, the adjustable upper and lower limits of the i-th scheduling unit at time t); scheduling response time (i.e., the time required for the ith scheduling unit to accept scheduling); responsive time period (this part of information can be incorporated into the prediction information, i.e., within this time period, the adjustable upper and lower limits are both 0); response rate (i.e., the ramp rate of the i-th dispatch unit); dispatch cost (i.e., the response cost of the i-th dispatch unit), response uncertainty (Initially 0, updated according to historical scheduling conditions, if multiple scheduling responses are poor, the value will be increased.) The following is an example:

[0107] Table 2 Example of dispatchable information of virtual power plant

[0108]

[0109] Table 3 Dispatch capacity prediction information of each dispatch unit

[0110]

[0111]

[0112] In addition to the common parameters of the above table equipment, it also includes some equipment-specific information, such as the installed capacity V of the energy storage unit i =500, and the current load state SOC i (t).

[0113] Step 4, real-time dispatch and adjustment. In the regional power market, the system real-time dispatch and adjustment mechanism closely integrates the electric energy market with the ancillary service market to ensure the safe and stable operation of the power grid and maximize economic benefits. In the pre-clearing stage of the day ahead, this mechanism follows the principle of "calling on demand and calling in sequence", sorts according to the unit price of deep peak regulation, and gives priority to calling units with low cost and fast response speed, providing strong support for real-time dispatch during the day. When encountering emergencies such as insufficient system margin, through the analysis of rolling unit combination results, the setting of must-start / must-stop units is started in time, and additional market clearing processes are run to call standby units from the ancillary service market to quickly respond to changes in power grid demand.

[0114] In the real-time operation process, the system dynamically adjusts the output of each unit according to the energy output demand and the market transaction settlement process to achieve optimal resource allocation. On-site transactions are settled based on the difference contract to ensure the rights and interests of both parties to the contract; while off-site transactions flexibly handle the financial impact of market fluctuations by channeling unbalanced funds. This series of real-time dispatch and adjustment measures together constitute the core of the economic dispatch decision-making plan in the region, effectively ensuring the stable operation of the power grid and the healthy development of the power market.

Claims

1. A multi-virtual power plant intelligent scheduling decision method considering uncertain risks, characterized in that: The following steps are involved: Construct an objective function to comprehensively optimize the responsiveness, total response cost, and uncertainty risk of provincial dispatching demand; set constraints in the intelligent dispatching decision model of the virtual power plant, including setting upper and lower limits of the response power, setting response rate constraints, setting response time constraints, and setting load state constraints of the energy storage system; data collection and preprocessing, the virtual power plant collects diversified data through user-level measurement units, and the decision-making layer preprocesses the data; Real-time scheduling and adjustment.

2. According to claim 1, a multi-virtual power plant intelligent scheduling decision method considering uncertain risks is characterized in that: The objective function is expressed as: minf=ω1f response +ω2f cost +ω3f risk In the formula, ω1, ω2, ω3 represent weight coefficients, f response 、f cost 、f risk They represent the objective functions of provincial dispatch demand satisfaction, dispatch cost and dispatch risk respectively.

3. The intelligent dispatching decision method for multiple virtual power plants considering uncertain risks according to claim 2 is characterized in that: The total response cost means that when there are multiple resources in the power dispatching area that can respond to the provincial dispatching, the dispatching units with lower dispatching costs should be given priority to ensure that the total dispatching response cost is minimized, that is, minf cost , In the formula, t start is the scheduling start time, t end The scheduling end time; Represents the unit cost of scheduling response of the i-th resource.

4. The method for intelligent scheduling and decision-making of multiple virtual power plants considering uncertain risks according to claim 2 is characterized in that: The uncertainty risk refers to the screening of resources according to the historical response of dispatchable resources in the dispatch decision process, giving priority to dispatching resources with high response stability, introducing dispatch risk cost, and ensuring the minimum risk cost, that is, minf risk , In the formula, C risk represents the unit energy scheduling risk cost, Represents the response uncertainty of the i-th scheduling resource, which is 0 initially.

5. The method for intelligent scheduling and decision-making of multiple virtual power plants considering uncertain risks according to claim 1 is characterized in that: The upper and lower limits of the response power are set to define the maximum power and minimum power range that the VPP can provide: Where p i (t) Represents the corresponding power, adjustable lower limit and upper limit of the i-th scheduling unit at time t.

6. The intelligent scheduling decision method for multiple virtual power plants considering uncertain risks according to claim 5 is characterized in that: The response time constraint is set as: t start ≤t≤t end In the formula, t start Indicates the time when the scheduling demand starts, t end Indicates the time when the scheduling demand ends, and t0 indicates the time when the scheduling demand instruction is issued; Indicates predicted output data.

7. The method for intelligent scheduling and decision-making of multiple virtual power plants considering uncertain risks according to claim 5 is characterized in that: The load state constraint of the energy storage system is set as: 0≤SOC i (t)*V i -p i (t)≤V i When the i-th dispatch resource is energy storage, the energy storage state of charge constraint should also be met. i (t) represents the state of charge of the energy storage at time t, V i Represents the installed capacity of energy storage.

8. The method for intelligent scheduling and decision-making of multiple virtual power plants considering uncertain risks according to claim 1 is characterized in that: The real-time dispatch and adjustment are specifically as follows: in the virtual power plant intelligent dispatch system, the regional ancillary service market and the electric energy market are integrated, and the day-ahead pre-clearing is based on the principles of demand and sequence, and is sorted according to the unit price of deep peak regulation for intra-day use.

9. The method for intelligent scheduling decision of multiple virtual power plants considering uncertain risks according to claim 8 is characterized in that: In the real-time scheduling and adjustment, the system dynamically adjusts the output of each unit according to the energy output demand and the market transaction settlement process to achieve optimal allocation of resources.

10. A multi-virtual power plant intelligent dispatching decision system considering uncertain risks, characterized in that: include: Objective function building module: When building an intelligent dispatching decision model for the collaboration between the virtual power plant and the regional power grid dispatching center, its core objective function aims to comprehensively optimize: the degree of response to provincial dispatching demand, the total response cost, and the uncertainty risk; Constraint setting module in the intelligent dispatch decision model of virtual power plants: When building the intelligent dispatch decision model of virtual power plants, a series of constraints need to be strictly set; Data collection and preprocessing module: The virtual power plant collects diverse data through user-level measurement units as the basis for sensing, predicting and controlling distributed generation and controllable resources; Real-time dispatching and adjustment module: In the intelligent dispatching system of the virtual power plant, the regional ancillary service market is integrated with the electric energy market, and the day-ahead pre-clearing is based on the principle of demand and sequence, and is sorted according to the unit price of deep peak regulation for intraday use.

Images