Virtual power plant intelligent scheduling optimization method

Abstract

The invention relates to the technical field of power systems and energy management, and discloses an intelligent scheduling optimization method for a virtual power plant, which comprises the following steps: constructing a multi-level scheduling model of the virtual power plant, and defining a multi-level scheduling architecture of the virtual power plant in the model, the framework comprises a distributed energy unit, a load subsystem, a distributed energy subsystem, an energy storage system and an aggregation load device module, the conditions of resource conflicts and bottlenecks occurring in the scheduling process of the virtual power plant are predicted and identified by using the lambda number, the scheduling strategy of the virtual power plant is dynamically adjusted according to the prediction result of the lambda number, and the scheduling efficiency of the virtual power plant is improved. According to the multi-level power plant resource collaborative optimization method, by introducing the lambda number in the scheduling process, bottlenecks and conflicts in the scheduling process are recognized and avoided, the risk values of different conflicts are analyzed based on the online value, and the power plant resource collaborative optimization method is suitable for the collaborative optimization of the scheduling targets of all levels in the multi-level optimization process. And intervention measures are adopted to reduce conflicts with high-risk value, so that optimal scheduling in a global range is realized.

Description

Technical Field

[0001] The present invention relates to the technical field of power system and energy management, and in particular to a method for intelligent dispatching and optimizing a virtual power plant. Background Art

[0002] As an important part of the energy internet, virtual power plants integrate distributed energy resources to achieve flexible and efficient energy scheduling and management. However, with the increase in the number of distributed energy accesses, the scheduling complexity of virtual power plants has increased significantly. In traditional virtual power plant demand scheduling, due to the complexity and dynamic changes of the system, the scheduling strategy is often difficult to achieve optimality on a global scale, and bottlenecks and resource conflicts are prone to occur. For example, in the peak shaving response of virtual power plants, there is a risk of conflict in user responses (some users actively participate in the response, some invalid responses or even reverse participation, making the total response of the virtual power plant invalid). To solve this problem, the present invention proposes a virtual power plant intelligent scheduling optimization method based on Ramsey numbers, introduces the Ramsey number theory in combinatorial mathematics, predicts and manages the inevitable risks in the scheduling process, and analyzes the reliability metric corresponding to the Ramsey number from the risk value theory VAR, thereby improving the scheduling efficiency and robustness of the system. Summary of the invention

[0003] 1. Technical issues to be solved

[0004] In view of the shortcomings of the prior art, the present invention provides a virtual power plant intelligent scheduling optimization method, which has the advantages of high scheduling efficiency and robustness. It solves the problem that in traditional virtual power plant demand scheduling, due to the complexity and dynamic changes of the system, the scheduling strategy is often difficult to achieve the best on a global scale, and bottlenecks and resource conflicts are prone to occur.

[0005] (II) Technical solution

[0006] To achieve the above object, the present invention provides the following technical solution: a virtual power plant intelligent scheduling optimization method, comprising the following steps:

[0007] Step 1: Establishment of model and framework: Construct a multi-level dispatch model of virtual power plant, define the multi-level dispatch architecture of virtual power plant in the model, the framework includes distributed energy units, load subsystems, distributed energy subsystems, energy storage systems and aggregated load modules, and list the relationships between the frameworks;

[0008] Step 2: Introduce prediction and identification technology: Introduce Ramsey number technology to predict bottlenecks, analyze resource conflicts and bottlenecks that will occur during virtual power plant scheduling, and use Ramsey number to predict and identify the above unavoidable situations;

[0009] Step 3: Dynamically adjust the scheduling strategy: According to the prediction result of the Ramsey number in step 2, the scheduling strategy is dynamically adjusted based on the risk value analysis of different conflict severities;

[0010] Step 4: Optimize resource allocation and improve robustness: Use the Ramsey number analysis results to optimize the spatial and temporal allocation of power plant resources;

[0011] Step 5: Multi-level intelligent scheduling optimization: In the multi-level optimization process, the Ramsey number technology is combined to coordinately optimize the scheduling objectives of each level.

[0012] Preferably, in step one, the virtual power plant is divided into multiple levels according to resource levels, and the levels include subsystem layer, resource layer, aggregation layer and equipment layer, specifically including distributed energy units, energy storage systems and aggregated loads.

[0013] Preferably, the distributed energy unit integrates all distributed energy within the virtual power plant, including wind power generation, solar power generation and small hydropower generation. Each unit is regarded as a module, and its power generation capacity, power generation cost, operation and maintenance cost, load volatility, and connection status with the external power grid are recorded.

[0014] Preferably, the energy storage system records the capacity, efficiency, charge and discharge rate and charge and discharge power parameters of the energy storage device, as well as the charge and discharge prices in different time periods, and the state of the energy storage system must be dynamically updated to reflect the real-time power level.

[0015] Preferably, the load aggregation module integrates the user loads in the virtual power plant, analyzes and predicts load demand fluctuations, and by collecting historical load data of each user in the virtual power plant, the load aggregation module establishes a demand model based on the historical load data to predict load demand in future time periods.

[0016] Preferably, in step 2, Ramsey number technology is introduced to predict bottlenecks, and the prediction process is: analyze historical data, analyze the corresponding bottlenecks, analyze the volatility index of returns occurring at different bottlenecks, and calculate the corresponding return standard deviation on this basis, and on this basis calculate the risk value of each scenario based on the following formula, which is:

[0017] VaR=Za*sigama*sqrt(T)

[0018] In the formula, VaR refers to the maximum possible loss of a financial instrument in a specific period of time in the future under a certain confidence level, Za represents the a quantile of the normal distribution, sigama represents the scenario volatility index, and T represents the length of time considered, which is one day, that is, T = 1. In financial risk management, the time period T is directly proportional to the VaR value;

[0019] Reorder the bottlenecks based on the VaR value. The larger the VaR, the greater the risk.

[0020] Preferably, the dynamic adjustment of the scheduling strategy in step three includes scheduling strategy design, real-time optimization and feedback mechanism. The scheduling strategy design adjusts the scheduling strategy of each unit in the virtual power plant according to the result of bottleneck identification. The real-time optimization and feedback mechanism tracks the status of each unit in the virtual power plant and the load level of the entire system by establishing a real-time monitoring system.

[0021] Preferably, the optimization configuration in step 4 includes: analyzing the impact of external power grid fluctuations, market price changes and load demand fluctuations on the system.

[0022] Preferably, the multi-level intelligent scheduling optimization in step five includes: coordinating scheduling strategies between levels under a multi-level optimization framework.

[0023] Preferably, the multi-level intelligent scheduling optimization in step five: collecting task requirements, resource status and environmental condition data of the power plant, developing an intelligent scheduling system based on the data of the power plant, integrating a Ramsey number calculation module, a multi-objective optimization algorithm and a feedback mechanism.

[0024] Compared with the prior art, the present invention provides a virtual power plant intelligent scheduling optimization method, which has the following beneficial effects:

[0025] 1. The introduction of Ramsey number technology in the present invention enables the system to predict and avoid possible resource conflicts and bottlenecks before scheduling. It can analyze historical data through calculation formulas. According to the numerical range of VaR value, when the VaR value is greater than the expected value, the virtual power plant model will take positive measures to avoid scheduling failures caused by system complexity. Through the bottleneck analysis strategy combining qualitative and quantitative methods, it can not only enhance the system's ability to cope with uncertainty and emergencies, but also ensure the stable operation of the virtual power plant in various complex environments, thereby improving the robustness of the model. BRIEF DESCRIPTION OF THE DRAWINGS

[0026] Figure 1 It is a flow chart of the steps of the present invention. DETAILED DESCRIPTION

[0027] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0028] See also Figure 1, a virtual power plant intelligent scheduling optimization method, comprising the following steps:

[0029] Step 1: Establishment of model and framework: Construct a multi-level dispatch model of virtual power plant, define the multi-level dispatch architecture of virtual power plant in the model, the framework includes distributed energy units, load subsystems, distributed energy subsystems, energy storage systems and aggregated load modules, and list the relationships between the frameworks;

[0030] Step 2: Introduce prediction and identification technology: Introduce Ramsey number technology to predict bottlenecks, analyze resource conflicts and bottlenecks that will occur during virtual power plant scheduling, and use Ramsey number to predict and identify the above unavoidable situations;

[0031] Step 3: Dynamically adjust the scheduling strategy: According to the prediction results of the Ramsey number, the scheduling strategy is dynamically adjusted based on the risk value analysis of different conflict severities to optimize resource allocation to avoid potential bottlenecks;

[0032] Step 4: Optimize resource allocation and improve robustness: Use the Ramsey number analysis results to optimize the spatial and temporal allocation of power plant resources and improve the robustness of the virtual power plant in complex environments;

[0033] Step 5: Multi-level intelligent scheduling optimization: In the multi-level optimization process, the Ramsey number technology is combined to coordinate the scheduling objectives of each level to ensure coordination and global optimization among levels.

[0034] In step 1, the virtual power plant is divided into multiple levels according to the resource hierarchy, which include subsystem layer, resource layer, aggregation layer and equipment layer, specifically including distributed energy units, energy storage systems and aggregated loads.

[0035] The distributed energy unit integrates all distributed energy within the virtual power plant, including wind power generation, solar power generation and small hydropower generation. Each unit is regarded as a module, recording its power generation capacity, power generation cost, operation and maintenance cost, load volatility, and connection status with the external power grid.

[0036] The energy storage system records the capacity, efficiency, charge and discharge rate and charge and discharge power parameters of the energy storage equipment, as well as the charge and discharge prices at different time periods. The status of the energy storage system must be dynamically updated to reflect the real-time power level.

[0037] Aggregate Load The load aggregation module integrates the user loads within the virtual power plant, analyzes and predicts load demand fluctuations. By collecting historical load data of each user in the virtual power plant, the load aggregation module will establish a demand model based on the historical load data to predict the load demand in future time periods.

[0038] In step 2, the Ramsey number technology is introduced to predict bottlenecks. The prediction process is as follows:

[0039] The Ramsey number describes the inevitable characteristics in a specific situation. In virtual power plant scheduling, the Ramsey number can help identify certain resource conflicts or bottlenecks in the scheduling process. These problems cannot be avoided by simple adjustments. At the same time, each edge in the system is analyzed to predict the conflicts that may arise under different power demand and supply situations. The following are some examples of prediction:

[0040] (1) When more than two distributed energy units supply power to an energy storage device at the same time and the total power supply exceeds the maximum allowable power of the device, a bottleneck may occur in the system.

[0041] (2) Based on the resource distribution and demand model, the possible Ramsey numbers and probabilistic Ramsey numbers that may occur in the system in different time periods are calculated. These values ​​represent the bottlenecks or resource conflicts that may occur in each time period.

[0042] According to the calculation results of the Ramsey number in embodiments (1) and (2), the bottlenecks that may occur in the scheduling are identified and classified into high, medium and low risk bottlenecks according to the size relationship of the Ramsey number. High-risk bottlenecks are usually those that appear repeatedly in multiple scenarios and need to be handled with priority. First, a risk priority list is established to guide subsequent scheduling strategy adjustments. In order to better quantify the risk level in different scenarios, determine the priority of strategy adjustment, and consider the online value in different scenarios, the scenario construction process is as follows:

[0043] S1. First, it is necessary to clarify the impact of bottlenecks on the overall system performance. The performance impact can be quantified from the following aspects:

[0044] S1.1. Economic impact: direct economic losses caused by bottlenecks. For example, certain bottlenecks may result in energy storage equipment not being fully utilized, thereby reducing the overall benefits of the system.

[0045] S1.2. Impact on system stability: Bottlenecks may lead to decreased system stability, such as frequency fluctuations, voltage instability, etc.

[0046] S1.3、Load supply and demand matching: Bottlenecks may lead to an imbalance between load and supply. It can be quantified by calculating the deviation between load demand and power supply capacity.

[0047] S1.4 Resource utilization efficiency: The utilization efficiency of distributed energy units and energy storage systems may be reduced. For example, some resources may be idle or unable to function fully due to bottlenecks.

[0048] S2. Analyze historical data, analyze the corresponding bottlenecks, analyze the return volatility index sigama of different bottlenecks in S1, and calculate the corresponding return standard deviation on this basis. On this basis, calculate the risk value of each scenario based on the following formula:

[0049] VaR=Za*sigama*sqrt(T)

[0050] In the formula, VaR refers to the maximum possible loss of a financial instrument in a specific period of time in the future under a certain confidence level. By comparing the VaR values ​​of different assets or projects, their risk sizes can be evaluated and ranked. Za represents the a quantile of the normal distribution. Sigma represents the scenario volatility index, which refers to the standard deviation of the asset return rate, reflects the volatility of the asset price, and is used to measure the potential range of asset price changes. T represents the length of time considered, which is one day, that is, T = 1. In financial risk management, the time period T is directly proportional to the VaR value, because the risk increases with the extension of time.

[0051] S3. Reorder the bottlenecks based on VaR values. The larger the VaR, the greater the risk.

[0052] The advantages are: the introduction of Ramsey number technology in the present invention enables the system to predict and avoid possible resource conflicts and bottlenecks before scheduling, and can analyze historical data through calculation formulas. According to the numerical range of VaR value, when the VaR value is greater than the expected value, the virtual power plant model will take positive measures to avoid scheduling failures caused by system complexity. Through the bottleneck analysis strategy that combines qualitative and quantitative methods, it can not only enhance the system's ability to cope with uncertainty and emergencies, but also ensure the stable operation of the virtual power plant in various complex environments, thereby improving the robustness of the model.

[0053] The dynamic adjustment of the dispatching strategy in step 3 includes dispatching strategy design, real-time optimization and feedback mechanism. The dispatching strategy design adjusts the dispatching strategy of each unit in the virtual power plant according to the results of bottleneck identification. The process is as follows:

[0054] First, high-risk bottlenecks are prioritized and emergency strategies are formulated. The goal of strategy adjustment is to reduce the probability of bottlenecks occurring, or to minimize the impact on the overall system when bottlenecks occur. Strategies are then adjusted, including adjusting the power generation plans of distributed energy units, changing the charging and discharging time of energy storage systems, allocating the needs of load aggregators, and reconfiguring power flow paths. By dynamically adjusting the scheduling strategy, the optimal allocation of resources can be achieved, thereby improving the utilization efficiency of distributed energy and energy storage systems.

[0055] Based on the scheduling design that combines multi-level distributed optimization with local adaptive control, local adaptive control modules are introduced for each distributed energy unit and energy storage system within the virtual power plant. Each module can independently adjust its own operating status according to real-time data and does not rely on the global scheduling strategy in the event of an emergency. This adaptive control enables the system to respond quickly to local changes, such as load fluctuations or equipment failures in a short period of time, thereby improving the stability and reaction speed of the system. Combined with local adaptive control, a multi-level distributed optimization algorithm is developed to decompose the global scheduling into multiple sub-problems, which are solved independently by local modules and then coordinated and integrated at the global scheduling level to speed up the calculation and reduce the burden on the central scheduling system.

[0056] Real-time optimization and feedback mechanism, by establishing a real-time monitoring system to track the status of each unit in the virtual power plant and the load level of the entire system, adjust the scheduling strategy in real time based on the monitoring data, introduce a feedback mechanism, and continuously update the calculation of the Ramsey number and bottleneck prediction during the scheduling process to ensure that the dynamic adjustment of the scheduling strategy can respond to system changes in a timely manner.

[0057] The optimization configuration in step 4 includes: analyzing the impact of external grid fluctuations, market price changes and load demand fluctuations on the system, using Ramsey number technology to identify the system's vulnerabilities under uncertain conditions, and formulating robustness improvement strategies, introducing necessary redundant designs into the system, such as backup power supplies and energy storage capacity, to ensure that the system can maintain stable operation in emergency situations.

[0058] The multi-level intelligent scheduling optimization in step five includes: under the multi-level optimization framework, coordinating the scheduling strategies between levels to ensure the coordination and consistency of the overall scheduling. The optimization results of each level will be fed back to the global scheduling strategy and iteratively adjusted. Through multi-level intelligent scheduling optimization, the resource allocation between levels is more coordinated, which helps to achieve the global optimal scheduling of the system.

[0059] Multi-level intelligent scheduling optimization in step five: collect task requirements, resource status and environmental conditions data of the power plant (data is the basis for subsequent decision-making analysis), develop an intelligent scheduling system based on the power plant data (the system contains an intelligent scheduling architecture, which automatically assigns tasks to appropriate resources and continuously optimizes scheduling plans to improve efficiency and response speed), integrate the Ramsey number calculation module (the Ramsey number calculation module is used to evaluate the collaborative efficiency between various resources in the virtual power plant and help determine the optimal resource allocation plan. By calculating the Ramsey numbers of different resource combinations, the system can identify which resource combinations can provide higher flexibility and stability), multi-objective optimization algorithms and feedback mechanisms to achieve automated scheduling optimization and ensure efficient operation of the system in a complex environment.

[0060] Although embodiments of the present invention have been shown and described, it will be appreciated by those skilled in the art that various changes, modifications, substitutions and variations may be made to the embodiments without departing from the principles and spirit of the invention, and that the scope of the invention is defined by the appended claims and their equivalents.

Claims

1. A virtual power plant intelligent scheduling optimization method, characterized in that: The following steps are involved: Step 1: Establishment of model and framework: Construct a multi-level dispatch model of virtual power plant, define the multi-level dispatch architecture of virtual power plant in the model, the framework includes distributed energy units, load subsystems, distributed energy subsystems, energy storage systems and aggregate load modules, and list the relationships between the frameworks; Step 2: Introduce prediction and identification technology: Introduce Ramsey number technology to predict bottlenecks, analyze resource conflicts and bottlenecks that will occur during virtual power plant scheduling, and use Ramsey number to predict and identify the above unavoidable situations; Step 3: Dynamically adjust the scheduling strategy: According to the prediction result of the Ramsey number in step 2, the scheduling strategy is dynamically adjusted based on the risk value analysis of different conflict severities; Step 4: Optimize configuration: Use the Ramsey number analysis results to optimize the spatial and temporal allocation of power plant resources; Step 5: Multi-level intelligent scheduling optimization: In the multi-level optimization process, the Ramsey number technology is combined to coordinately optimize the scheduling objectives of each level.

2. The method for intelligent scheduling and optimization of a virtual power plant according to claim 1, characterized in that: In step one, the virtual power plant is divided into multiple levels according to the resource hierarchy, and the levels include subsystem layer, resource layer, aggregation layer and equipment layer, specifically including distributed energy units, energy storage systems and aggregated loads.

3. A virtual power plant intelligent scheduling optimization method according to claim 2, characterized in that: The distributed energy unit integrates all distributed energy within the virtual power plant, including wind power generation, solar power generation and small hydropower generation. Each unit is regarded as a module, recording its power generation capacity, power generation cost, operation and maintenance cost, load volatility, and connection status with the external power grid.

4. The method for intelligent scheduling and optimization of a virtual power plant according to claim 2, characterized in that: The energy storage system records the capacity, efficiency, charge and discharge rate and charge and discharge power parameters of the energy storage equipment, as well as the charge and discharge prices in different time periods. The status of the energy storage system must be dynamically updated to reflect the real-time power level.

5. The method for intelligent scheduling and optimization of a virtual power plant according to claim 2, characterized in that: The load aggregation module integrates the user loads in the virtual power plant, analyzes and predicts load demand fluctuations, and by collecting historical load data of each user in the virtual power plant, the load aggregation module establishes a demand model based on the historical load data to predict load demand in future time periods.

6. The method for intelligent scheduling and optimization of a virtual power plant according to claim 1, characterized in that: In step 2, the Ramsey number technology is introduced to predict bottlenecks. The prediction process is as follows: analyze historical data, analyze the corresponding bottlenecks, analyze the volatility index of returns caused by different bottlenecks, and calculate the corresponding return standard deviation on this basis. On this basis, the risk value of each scenario is calculated based on the following formula: VaR=Za*sigama*sqrt(T) In the formula, VaR refers to the maximum possible loss of a financial instrument in a specific period of time in the future under a certain confidence level, Za represents the a quantile of the normal distribution, sigama represents the scenario volatility index, and T represents the length of time considered, which is one day, that is, T = 1. In financial risk management, the time period T is directly proportional to the VaR value; Reorder the bottlenecks based on the VaR value. The larger the VaR, the greater the risk.

7. The method for intelligent scheduling and optimization of a virtual power plant according to claim 1, characterized in that: The dynamic adjustment of the scheduling strategy in step three includes scheduling strategy design, real-time optimization and feedback mechanism. The scheduling strategy design adjusts the scheduling strategy of each unit in the virtual power plant according to the results of bottleneck identification. The real-time optimization and feedback mechanism tracks the status of each unit in the virtual power plant and the load level of the entire system by establishing a real-time monitoring system.

8. The method for intelligent scheduling and optimization of a virtual power plant according to claim 1, characterized in that: The optimization configuration in step 4 includes: analyzing the impact of external power grid fluctuations, market price changes and load demand fluctuations on the system.

9. The method for intelligent scheduling and optimization of a virtual power plant according to claim 1, characterized in that: The multi-level intelligent scheduling optimization in step 5 includes: coordinating scheduling strategies between levels under a multi-level optimization framework.

10. A virtual power plant intelligent dispatch optimization method according to claim 9, characterized in that: The multi-level intelligent scheduling optimization in step five: collecting task requirements, resource status and environmental condition data of the power plant, developing an intelligent scheduling system based on the data of the power plant, integrating the Ramsey number calculation module, multi-objective optimization algorithm and feedback mechanism.

Images