Virtual power plant source load storage power optimization scheduling method and system
By building a virtual power plant operation model and adopting an improved cuckoo search algorithm, combined with the demand-side response mechanism, the limitations of the optimization scheduling algorithm and insufficient adaptability of renewable energy volatility in virtual power plants are solved, and a more efficient economy is achieved Optimization of performance and environmental protection has improved the utilization efficiency of energy storage systems.
Patent Information
- Application Number
- CN202410820551.6
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-06-24
- Publication Date
- 2024-11-19
AI Technical Summary
The existing technology optimization scheduling algorithm in virtual power plants has problems such as slow convergence speed, easy to fall into local optimality, and complex parameter adjustment, lack of adaptability to renewable energy volatility, insufficient development of demand-side management potential, and energy storage system scheduling The strategy is single and it is difficult to comprehensively consider multi-objective scheduling models.
Build a virtual power plant operation model, including distributed power generation, energy storage and load unit models, use the improved cuckoo search algorithm for global and local search, combine the demand-side response mechanism, optimize the energy storage system scheduling, and establish comprehensive consideration of economic, environmental protection and Multi-objective optimization model for user satisfaction.
It improves the economic operation efficiency of virtual power plants and adapts to renewable energy volatility, realizes the optimization of demand-side management, improves the utilization efficiency and overall performance of energy storage systems, and solves key problems in optimized scheduling.
Smart Images

Figure CN118982160A_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to the technical field of virtual power plants, and in particular to a method and system for optimizing the dispatching of sources, loads and storage power of virtual power plants. Background Art
[0002] In today's world, with the increasing depletion of fossil energy and the increasingly serious environmental pollution problems, the traditional energy system is facing unprecedented challenges. In order to meet these challenges, the global energy structure is undergoing a profound transformation, that is, from dependence on fossil energy to renewable energy. In this transformation process, virtual power plants, as an important part of the new energy system, have received great attention for their development and optimized operation.
[0003] A virtual power plant is a small power network that includes distributed generation resources, energy storage systems, loads, and corresponding monitoring and control systems. It can operate in both grid-connected and island modes, which can not only improve the reliability and flexibility of the power system, but also effectively promote the use of renewable energy such as wind and solar energy. The optimal scheduling of virtual power plants is one of the key technologies to ensure their efficient operation. It needs to achieve the optimal matching of sources, loads, and storage in virtual power plants under the premise of meeting system operation constraints, battery operation constraints, and user satisfaction, so as to achieve the dual goals of economy and environmental protection.
[0004] For example, the existing invention patent application document "Virtual Power Plant Distributed Photovoltaic Energy Storage System Configuration Optimization Method and Device" with publication number CN118070988A, the existing method includes: according to the acquired virtual power plant operation status data, a multi-objective scheduling model with maximum economic benefits, minimum intraday power fluctuation and optimal global voltage quality evaluation is obtained; the multi-objective scheduling model of the virtual power plant is mapped to the multi-objective configuration optimization model of the energy storage system, and the single-objective configuration optimization model is obtained after normalizing the multi-objective configuration optimization model, and the single-objective configuration optimization model is solved to obtain the optimal configuration result. The aforementioned existing virtual power plant optimization scheduling research mainly focuses on the following aspects: first, how to accurately predict and schedule renewable energy generation in virtual power plants, especially the uncertainty of wind and solar power generation; second, how to reasonably schedule energy storage systems to balance load demand and power generation supply, especially during peak power demand periods; finally, how to optimize load distribution and reduce system operating costs through demand-side management under the premise of ensuring user satisfaction. In the prior art, a variety of optimization algorithms have been proposed to solve the optimization scheduling problem of virtual power plants. For example, genetic algorithms, particle swarm optimization algorithms, simulated annealing algorithms, etc. are all common optimization tools. These algorithms have their own advantages. For example, genetic algorithms excel in global search capabilities, particle swarm optimization algorithms are more efficient in dealing with continuous optimization problems, and simulated annealing algorithms are good at solving optimization problems with higher complexity. However, these algorithms still have some limitations when dealing with virtual power plant optimization scheduling problems. For example, they may be insufficient in terms of convergence speed, solution quality, and algorithm stability.
[0005] Another example is the existing invention patent application document "Virtual Power Plant Optimization Scheduling Method Considering Electric Vehicle Demand Response and Its Application" with publication number CN117543581A. The existing method includes: virtual power plant system structure, demand response EVG operation strategy, construction of electric vehicle dispatchable potential prediction model, virtual power plant scheduling modeling. The present invention designs operation strategies for controllable charging vehicles and uncontrollable charging vehicles, constructs two dispatchable potential prediction models for controllable charging vehicles and uncontrollable charging vehicles based on demand response operation strategies, inputs wind and solar power output and load forecast values, establishes objective functions and constraints, and solves the optimal electricity price based on the demand response EVG operation strategy using a particle swarm algorithm. In the field of virtual power plant source-load-storage power optimization scheduling, the aforementioned prior art has limitations, which not only affect the economic operation efficiency of virtual power plants, but also may limit their application potential in smart grids. The following are the shortcomings of the prior art and the technical problems that the present invention aims to solve:
[0006] Disadvantages of existing technology:
[0007] Limitations of optimization algorithms: Although traditional optimization algorithms, such as genetic algorithms and particle swarm optimization algorithms, have achieved certain success in solving complex optimization problems, they may face problems such as slow convergence, easy to fall into local optimality, and complex parameter adjustment in virtual power plant scheduling problems. These limitations restrict the application effect of the algorithm in real-time and dynamic environments.
[0008] Lack of adaptability to the volatility of renewable energy: As the share of renewable energy (such as wind and solar) in virtual power plants increases, existing technologies are insufficient in dealing with the volatility of these energy sources. The uncertainty and unpredictability of wind and solar energy pose challenges to the stable operation and power dispatch of virtual power plants.
[0009] Insufficient demand-side management: Existing technologies may not fully utilize the potential of demand-side management (DSM), especially during peak hours, when strategies to reduce the burden on the grid through demand response and load management have not been fully developed and applied.
[0010] The singleness of the energy storage system dispatching strategy: The energy storage system plays a vital role in the virtual power plant, but the existing dispatching strategies often lack flexibility and intelligence and cannot be optimized and adjusted according to the actual needs and operating status of the power grid.
[0011] Lack of a scheduling model that comprehensively considers multiple objectives: The optimal scheduling of virtual power plants must not only consider economic benefits, but also environmental benefits, system stability, user satisfaction, etc. Existing technologies often find it difficult to find the best balance between these objectives.
[0012] The existing invention patent application document "A load-side virtual power plant scheduling method based on an improved swarm intelligence algorithm" with publication number CN117595239A, the existing method includes: establishing a central air-conditioning physical model; solving the chilled water parameters, fan coil parameters and room parameters of the central air-conditioning physical model by an enhanced sparrow search algorithm, and obtaining the functional relationship between the electric power and indoor temperature of the central air-conditioning physical model; optimizing the virtual power plant scheduling by a two-way raid multi-objective gray wolf algorithm. The present invention reduces the load of the central air-conditioning by an enhanced sparrow search algorithm under the conditions of satisfying temperature satisfaction and providing pricing compensation to users, and obtains the virtual power plant scheduling optimization strategy by a two-way raid multi-objective gray wolf algorithm. However, the improved swarm intelligence algorithm proposed in the above-mentioned prior art may have limitations in multi-objective scheduling.
[0013] The existing invention patent application document "A virtual power plant optimization scheduling method based on dual-mode economic model predictive control" with publication number CN117151357A, the existing method includes: first designing the virtual power plant optimization scheduling objective function and constraints including wind power, photovoltaic, gas turbine, battery energy storage and interconnection line, and solving the steady-state economic optimal point; then, according to the virtual power plant operation constraints, designing the state space prediction model of economic model predictive control; finally, taking the steady-state economic optimal point as the tracking target, designing the virtual power plant optimization operation control strategy based on dual-mode economic model predictive control, mode 1 improves the operation economy through economic model predictive control, and mode 2 maintains the tracking performance with the help of the designed auxiliary controller. However, the two-stage coordinated optimization scheduling method in the above-mentioned prior art may be insufficient in dynamic tracking performance.
[0014] The existing invention patent application document with publication number "CN114037191A Virtual power plant optimization scheduling method, device, equipment and medium based on big data" includes: taking the virtual power plant to maximize profits as the goal, constructing the objective function of the economic scheduling of the virtual power plant, and combining the objective function with the operating constraints of the virtual power plant to generate the day-ahead scheduling plan for each virtual power plant corresponding to the joint system; determining the wind power forecast output corresponding to the intraday scheduling stage according to the day-ahead scheduling plan, comparing the wind power forecast output with the actual wind power output of the control domain step, and obtaining the output forecast error of the control domain step; iteratively optimizing and adjusting the step time of the prediction domain step according to the output forecast error, and iteratively optimizing and correcting the control variables corresponding to the day-ahead scheduling plan. However, the optimization method based on big data in the aforementioned prior art may lack adaptability to the volatility of renewable energy.
[0015] In summary, the existing technologies have technical problems such as limited optimization algorithms, lack of adaptability to the volatility of renewable energy, insufficient demand-side management, single scheduling strategy of energy storage systems, and lack of scheduling models that comprehensively consider multiple objectives. Summary of the invention
[0016] The technical problem to be solved by the present invention is: how to solve the technical problems in the prior art such as the limitation of optimization algorithms, lack of adaptability to the volatility of renewable energy, insufficient demand-side management, the singleness of the energy storage system scheduling strategy, and the lack of a scheduling model that comprehensively considers multiple objectives.
[0017] The present invention adopts the following technical solutions to solve the above technical problems: the virtual power plant source-load-storage power optimization scheduling method comprises:
[0018] S1. Construct a virtual power plant operation model, including: a distributed power generation unit model, an energy storage unit model and a load unit model; in the virtual battery operation model, the user loads that meet the power supply model are divided into fixed loads, movable loads and disconnectable loads for scheduling optimization, wherein the distributed power generation unit model includes: a wind power generation model and a photovoltaic power generation model;
[0019] S2. Construct a virtual battery optimization scheduling model, including economic objective function and environmental objective function, and set constraints, including system operation constraints, battery operation constraints and user satisfaction constraints; system operation constraints include power balance equation constraints and grid interaction power constraints; battery operation constraints include battery charging and discharging constraints, battery state constraints and battery initial and final energy state constraints;
[0020] S3. Adjust and obtain an applicable cuckoo search algorithm CS to solve the virtual power plant operation model and the virtual battery optimization scheduling model; wherein, using the applicable cuckoo search algorithm CS, the cuckoo parasitic behavior is simulated and Levy flight is performed, global search and local search are performed, and the bird's nest solution position is updated to obtain the optimal solution of the model;
[0021] S4. Run the applicable cuckoo search algorithm CS to obtain the economic optimization scheduling plan of the virtual power plant based on the optimal solution of the model.
[0022] The present invention constructs a multi-objective optimization scheduling model that comprehensively considers economic benefits, environmental benefits, system stability and user satisfaction. Through the improved cuckoo algorithm, the optimal scheduling of the source-load-storage power of the virtual power plant under multiple constraints is achieved.
[0023] The present invention proposes a more flexible and intelligent energy storage system scheduling strategy, which can be optimized and adjusted according to the actual needs and operating status of the power grid, improving the utilization efficiency of the energy storage system and the overall performance of the virtual power plant. The present invention not only solves some key problems in the optimization and scheduling of virtual power plants, but also improves the solution efficiency and solution quality by improving the algorithm. This research result has important theoretical and practical significance for promoting the development and application of virtual power plant technology.
[0024] In a more specific technical solution, S1 includes:
[0025] S11. Construct a wind power generation model according to the following wind power output determination logic:
[0026]
[0027] Where P wt is the fan output power, ρ is the air density, R is the rotation radius of the fan blade, ν is the maximum linear velocity on the fan blade, Cp is the wind power conversion coefficient;
[0028] S12. Construct a photovoltaic power generation model according to the following photovoltaic output power determination logic:
[0029]
[0030] Where P pv , L, T pv are the output power, light intensity and temperature of photovoltaic power generation under the target environment, P STC , L STC , T STC are the output power, light intensity and temperature of photovoltaic power generation under standard test conditions, k T is the power temperature coefficient;
[0031] S13. Construct an energy storage unit model using the following state of charge SOC value determination logic:
[0032]
[0033] In the formula, S t is the SOC value of the battery during period t, P sb,t is the charging and discharging power of the energy storage system in the tth period, E b is the rated capacity of the battery, and Δt is the time period.
[0034] S14. Construct a load unit model using the following total load determination logic:
[0035] P L (t) = P B (t)+P M (t)+P D (t)
[0036] Where P L (t) is the total load power in the virtual power plant during period t; P B (t), P M (t) and P B (t) are the fixed load power, movable load power and disconnectable load power in period t respectively.
[0037] S15. Model the movable load using the following logic:
[0038]
[0039] Where P M_in,i (t) is the transferable load power from period i to period t; P M_out (t) is the load power that can be removed during period t.
[0040] The present invention introduces a demand-side response mechanism and utilizes the scheduling of movable loads and disconnectable loads to optimize power demand in different time periods, reduce the operating costs of virtual power plants, and improve the stability and reliability of the system.
[0041] In a more specific technical solution, S2 includes:
[0042] S21. Using the following virtual power plant economic cost logic, set the economic objective function:
[0043] minC mg =C wt +C pv +C sb +C grid +C D +C M
[0044] In the formula, C mg is the economic cost of the microgrid, C wt is the cost of wind power generation, C pv is the photovoltaic power generation cost, C sb is the battery operating cost, C grid is the cost of exchanging electricity between the virtual power plant and the main grid, C D Compensation for users’ power interruption, C M Dispatch and compensate for users' electricity load;
[0045] S22. Using the following logic for determining the environmental protection cost of a virtual power plant, set an environmental protection objective function:
[0046]
[0047] In the formula, C mp is the environmental cost, e is the carbon emission cost coefficient, and E is the marginal emission factor.
[0048] S23. Determine the power balance equation constraint using the following logic:
[0049] P wt +P pv +P sb +P grid -P L =0
[0050] S24. Determine the grid interaction power constraint using the following logic:
[0051] -P gs,max ≤P grid ≤P gb,max
[0052] Where P gs,max P is the upper limit of the power sold by the virtual power plant to the main grid, expressed as a positive number;gb,max The upper limit of the power that the power grid can purchase from the main grid, expressed as a positive number;
[0053] S25. Determine the dispatch load power constraint using the following logic to limit the load dispatch power:
[0054]
[0055] Where P M_out,nax , P D,max They are the upper limit of the power of the removed load and the upper limit of the power of the disconnectable load respectively;
[0056] S26. Determine the battery charge and discharge constraints using the following logic:
[0057]
[0058] Where P sout,max is the maximum discharge power, P sin,max is the maximum charging power;
[0059] S27. Determine the battery state constraint using the following logic:
[0060] S min t max
[0061] In the formula, S max , S min They are the upper and lower limits of the state of charge (SOC) of the energy storage system respectively.
[0062] S28. Determine the user satisfaction constraint using the following logic:
[0063] M=U com U eco
[0064] In the formula, M is user satisfaction, U is com After optimizing for user load, U eco The economic degree of the user's electricity load;
[0065] S29, obtaining and calculating the user comfort level based on the difference between the user's actual power consumption curve and the maximum comfort level curve using the following logic:
[0066]
[0067] S210, according to the difference in the cost of electricity used by the user before and after load optimization, the user's economic degree is calculated using the following logic:
[0068]
[0069] In the formula, Cbef is the user's electricity cost before load dispatch, C aft The electricity cost of users after load scheduling.
[0070] The present invention constructs a virtual power plant active power dispatching model that comprehensively considers system operation constraints, battery operation constraints and user satisfaction. The improved cuckoo algorithm of this study can effectively deal with the source-load-storage coordination problem in the virtual power plant.
[0071] In the model of the present invention, we not only consider the uncertainty of wind and photovoltaic power generation, but also the dynamic changes of load and the operating limitations of energy storage equipment. Through application and verification in multiple typical scenarios, the results show that this method can effectively reduce the operating cost of virtual power plants while improving the environmental performance of the system.
[0072] In a more specific technical solution, in S21, the following logic is used to determine the economic cost of the microgrid, the cost of wind power generation, the cost of photovoltaic power generation, the operating cost of the battery, the cost of exchanging power between the virtual power plant and the main grid, the compensation for the user's power interruption, and the compensation for the user's power load dispatch:
[0073]
[0074] In the formula, α wt is the unit wind power generation cost; α pv is the unit photovoltaic power generation cost; R sout,t is the battery status during period t; β sb is the unit discharge cost of the battery; g B P is the unit electricity purchase cost of the microgrid during period t; grid is the interaction power between microgrid and main grid; D is the interruptible load compensation coefficient; a t,i Unit dispatch compensation for load shifting from period i to period t;
[0075] The unit dispatch compensation a for load shifting from period i to period t is determined by the following logic: t,i :
[0076]
[0077] Where a1, a2, a3, and a4 are the compensation coefficients of the movable load at different moving periods.
[0078] The present invention constructs a virtual power plant power dispatch model that includes wind and solar power generation, energy storage units and loads, comprehensively considers system operation constraints, battery operation constraints and user satisfaction, and takes the optimization of the economic and environmental performance of the virtual power plant as the objective function, thereby improving the adaptability of the virtual power plant to the volatility of renewable energy.
[0079] In a more specific technical solution, in S26, the following logic is used to determine the battery charge and discharge times condition:
[0080]
[0081] In the formula, R sin,t is the battery status during period t.
[0082] In a more specific technical solution, S3 includes:
[0083] S31. Perform a global search through Levy flight to update the Levy flight position using the following logic:
[0084]
[0085] In the formula, is the position of the jth (j=1,2,3,…,n) bird’s nest in the kth generation, is point-to-point multiplication, α is the step size control value, which is used to control the step size;
[0086] S32. Perform local search through local random walk:
[0087]
[0088] In the formula, and are the positions of two random nests in the contemporary nest, H(μ) is the Heaviside function, P a is the maximum discovery probability, s is the moving step length, and ε is a uniformly distributed random number.
[0089] S33, using the following logic, using the dynamically changing step size enhancement algorithm to adjust the step size:
[0090]
[0091] Where a, b, c are step size control quantities, and a>b>c; k is the current iteration number; N is the total iteration number.
[0092] In a more specific technical solution, in step S31, the step size is determined using the following logic:
[0093]
[0094] Γ(λ)=∫0 ∞ t λ-1 e -t dt.
[0095] The present invention improves the traditional cuckoo algorithm and introduces an adaptive step size strategy, so that the algorithm can dynamically adjust the search step size according to the current solution during the search process, thereby achieving a better balance between global search and local search.
[0096] The improved cuckoo algorithm of the present invention not only improves the quality of the solution when solving the virtual power plant optimization scheduling problem, but also outperforms other algorithms in terms of computational efficiency. The improved cuckoo algorithm is an effective tool for virtual power plant optimization scheduling, which can provide strong support for the economic and environmental optimization of virtual power plants.
[0097] In a more specific technical solution, in step S31, the step length is calculated by the Mantegna algorithm:
[0098]
[0099] The present invention improves the global search capability and convergence speed of the algorithm through adaptive step size adjustment and Levy flight strategy, thereby more effectively solving the optimization scheduling problem of the virtual power plant.
[0100] In a more specific technical solution, in step S31, μ and ν obey the following normal distribution:
[0101] and
[0102] and σ ν =1.
[0103] In a more specific technical solution, the virtual power plant source-load-storage power optimization dispatching system includes:
[0104] The power plant operation modeling module is used to build a virtual power plant operation model, including: distributed power generation unit model, energy storage unit model and load unit model; in the virtual battery operation model, the user load that meets the power model is divided into fixed load, movable load and disconnectable load for scheduling optimization, among which the distributed power generation unit model includes: wind power generation model and photovoltaic power generation model;
[0105] The battery scheduling modeling module is used to build a virtual battery optimization scheduling model, including economic objective functions, environmental objective functions, and set constraints, including: system operation constraints, battery operation constraints, and user satisfaction constraints; system operation constraints include: power balance equation constraints and grid interaction power constraints; among them, battery operation constraints include: battery charging and discharging constraints, battery state constraints, and battery initial and final energy state constraints;
[0106] The model solving module is used to adjust and obtain the applicable cuckoo search algorithm CS to solve the virtual power plant operation model and the virtual battery optimization scheduling model; wherein, the applicable cuckoo search algorithm CS is used to simulate the cuckoo parasitic behavior and perform Levy flight, perform global search and local search, and update the bird's nest solution position to obtain the optimal solution of the model. The model solving module is connected with the power plant operation modeling module and the battery scheduling modeling module;
[0107] The scheduling scheme determination module is used to run the applicable cuckoo search algorithm CS to obtain the economic optimization scheduling scheme of the virtual power plant according to the optimal solution of the model. The scheduling scheme determination module is connected with the model solving module.
[0108] Compared with the prior art, the present invention has the following advantages:
[0109] The present invention constructs a multi-objective optimization scheduling model that comprehensively considers economic benefits, environmental benefits, system stability and user satisfaction. Through the improved cuckoo algorithm, the optimal scheduling of the source-load-storage power of the virtual power plant under multiple constraints is achieved.
[0110] The present invention proposes a more flexible and intelligent energy storage system scheduling strategy, which can be optimized and adjusted according to the actual needs and operating status of the power grid, improving the utilization efficiency of the energy storage system and the overall performance of the virtual power plant. The present invention not only solves some key problems in the optimization and scheduling of virtual power plants, but also improves the solution efficiency and solution quality by improving the algorithm. This research result has important theoretical and practical significance for promoting the development and application of virtual power plant technology.
[0111] The present invention introduces a demand-side response mechanism and utilizes the scheduling of movable loads and disconnectable loads to optimize power demand in different time periods, reduce the operating costs of virtual power plants, and improve the stability and reliability of the system.
[0112] The present invention constructs a virtual power plant active power dispatching model that comprehensively considers system operation constraints, battery operation constraints and user satisfaction. The improved cuckoo algorithm of this study can effectively deal with the source-load-storage coordination problem in the virtual power plant.
[0113] In the model of the present invention, we not only consider the uncertainty of wind and photovoltaic power generation, but also the dynamic changes of load and the operating limitations of energy storage equipment. Through application and verification in multiple typical scenarios, the results show that this method can effectively reduce the operating cost of virtual power plants while improving the environmental performance of the system.
[0114] The present invention constructs a virtual power plant power dispatch model that includes wind and solar power generation, energy storage units and loads, comprehensively considers system operation constraints, battery operation constraints and user satisfaction, and takes the optimization of the economic and environmental performance of the virtual power plant as the objective function, thereby improving the adaptability of the virtual power plant to the volatility of renewable energy.
[0115] The present invention improves the traditional cuckoo algorithm and introduces an adaptive step size strategy, so that the algorithm can dynamically adjust the search step size according to the current solution during the search process, thereby achieving a better balance between global search and local search.
[0116] The improved cuckoo algorithm of the present invention not only improves the quality of the solution when solving the virtual power plant optimization scheduling problem, but also outperforms other algorithms in terms of computational efficiency. The improved cuckoo algorithm is an effective tool for virtual power plant optimization scheduling, which can provide strong support for the economic and environmental optimization of virtual power plants.
[0117] The present invention improves the global search capability and convergence speed of the algorithm through adaptive step size adjustment and Levy flight strategy, thereby more effectively solving the optimization scheduling problem of the virtual power plant.
[0118] The present invention not only improves the economic operation efficiency of virtual power plants, but also enhances their adaptability to the volatility of renewable energy, realizes the optimization of demand-side management, improves the scheduling efficiency of energy storage systems, and finds the best balance point among multiple objectives, providing strong technical support for the sustainable development of virtual power plants and the construction of smart grids.
[0119] The present invention solves the technical problems existing in the prior art, namely, the optimization algorithm is limited, lacks adaptability to the volatility of renewable energy, inadequate demand-side management, the singleness of the energy storage system scheduling strategy, and lacks a scheduling model that comprehensively considers multiple objectives. BRIEF DESCRIPTION OF THE DRAWINGS
[0120] Figure 1 A schematic diagram of basic steps of a method for optimizing source-load-storage power scheduling of a virtual power plant according to Embodiment 1 of the present invention;
[0121] Figure 2 This is a schematic diagram of the structure of a virtual power plant system according to Embodiment 1 of the present invention;
[0122] Figure 3 This is a schematic diagram of the specific steps of running the CS algorithm according to Example 1 of the present invention. DETAILED DESCRIPTION
[0123] In order to make the purpose, technical solution and advantages of the embodiments of the present invention clearer, the technical solution in the embodiments of the present invention will be clearly and completely described below in combination with the embodiments of the present invention. Obviously, the described embodiments are 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.
[0124] Example 1
[0125] like Figure 1 As shown, the virtual power plant source-load-storage power optimization scheduling method provided by the present invention includes the following basic steps:
[0126] S1. Construction of virtual power plant system model;
[0127] In this embodiment, it is necessary to construct an operation model of a virtual power plant, which includes distributed generation units, energy storage units, and load units. The model ignores power transmission losses and divides user loads into three categories: fixed loads, movable loads, and disconnectable loads, for scheduling optimization.
[0128] In this embodiment, the distributed power generation units include but are not limited to: wind power and photovoltaic power generation; the energy storage units include: batteries.
[0129] like Figure 2 As shown, in this embodiment, a virtual power plant operation model is constructed. In this embodiment, the virtual power plant mainly includes three parts: distributed generation units, energy storage units and load units, and is directly connected to the same bus. Due to the small geographical range and short lines, this paper ignores power transmission losses in constructing the virtual power plant scheduling model.
[0130] Wind power generation model
[0131] Wind power generation uses wind power to drive blades to rotate, and the generator connected to the blades outputs electrical energy to achieve the conversion of wind energy into electrical energy.
[0132] The output power is shown as:
[0133]
[0134] Where: P wt is the fan output power; ρ is the air density; R is the rotation radius of the fan blade; ν is the maximum linear velocity on the fan blade; C p is the wind power conversion coefficient.
[0135] In the process of constructing the photovoltaic power generation model of this embodiment, photovoltaic power generation is based on the conversion of light energy into electrical energy by semiconductor materials. The key factors affecting power generation are the light intensity and the temperature around the semiconductor material. The output power is shown in the following formula:
[0136]
[0137] Among them, P pv , L, T pv are the output power, light intensity and temperature of photovoltaic power generation under the target environment; P STC , L STC , T STC They are the output power, light intensity and temperature of photovoltaic power generation under standard test conditions; k T is the power temperature coefficient.
[0138] In the process of constructing the energy storage unit model of this embodiment, the energy of the energy storage unit in each period in the virtual power plant is changing, and the state of charge SOC value is used to represent the power stored in the battery in each period. The SOC calculation expression for period t is as follows:
[0139]
[0140] in,
[0141] S t is the SOC value of the battery during period t; P sb,t E is the charging and discharging power of the energy storage system (battery) in the tth period, which takes a negative value during charging and a positive value during discharging; b is the rated capacity of the battery; Δt is the time period.
[0142] In the process of constructing the load model of this embodiment, user loads are divided into three categories from the perspective of load operation scheduling: fixed loads, movable loads, and disconnectable loads. Fixed loads refer to loads whose usage time and power are relatively fixed and cannot be scheduled; movable loads refer to loads whose usage time and power can be shifted to other time periods as needed; disconnectable loads refer to loads that can be partially or completely interrupted as needed. In terms of microgrid optimization scheduling, virtual power plants achieve economic optimization of operating costs by scheduling movable loads.
[0143] In this embodiment, the total load of the virtual power plant during period t is the sum of the fixed load, the movable load and the disconnectable load, and its expression is:
[0144] P L (t) = P B (t)+P M (t)+P D (t)
[0145] Among them, P L(t) is the total load power in the virtual power plant during period t; P B (t), P M (t) and P B (t) are the fixed load power, movable load power and disconnectable load power in period t respectively.
[0146] In this embodiment, for the movable load, when the load is moved, both the amount of load transferred in the current period and the amount of load transferred out in the same period should be considered. The expression of the movable load in a certain period is:
[0147]
[0148] Among them, P M_in,i (t) is the transferable load power from period i to period t; P M_out (t) is the load power that can be removed during period t.
[0149] S2, establishment of optimization scheduling model;
[0150] In the virtual power plant optimization scheduling model of this embodiment, the objective function includes two aspects: economy and environmental protection. The economic objective function takes into account the power generation cost, energy storage operation cost, transaction cost with the main grid, and load response subsidy. The environmental objective function takes CO2 emissions as an indicator and takes into account the carbon emission cost of thermal power generation in the main grid.
[0151] In the process of constructing the virtual power plant day-ahead optimization scheduling model of this embodiment, objective functions are set, including but not limited to: an economic objective function and an environmental objective function.
[0152] In this embodiment, an economic objective function is set. The economic cost of the virtual power plant is mainly composed of the power generation of each micro-source, the energy storage operation cost, the transaction cost between the microgrid and the main grid, and the load response subsidy. The economic cost of the virtual power plant can be expressed as:
[0153] minC mg =C wt +C pv +C sb +C grid +C D +C M
[0154] Among them, C mg is the economic cost of the microgrid; C wt is the cost of wind power generation; C pv is the photovoltaic power generation cost; C sb is the battery operating cost; C grid C is the cost of exchanging electricity between the virtual power plant and the main grid; D Compensation for users’ power outages; C MProvide compensation for users' electricity load scheduling.
[0155]
[0156] Among them, α wt is the unit wind power generation cost; α pv is the unit photovoltaic power generation cost; R sout,t is the battery status during period t, 1 means the battery is in a discharging state, and 0 means the battery is in a neither-charging nor-discharging state; β sb is the unit discharge cost of the battery; g B is the unit electricity purchase cost of the microgrid during period t, which is 0 when selling electricity; g S is the unit electricity sales revenue of the microgrid during period t, which is 0 when purchasing electricity; P grid is the interaction power between the microgrid and the main grid, which is positive when the microgrid purchases electricity and negative when the microgrid sells electricity; D is the interruptible load compensation coefficient; a t,i Unit dispatch compensation for load shifting from period i to period t:
[0157]
[0158] Among them, a1, a2, a3, and a4 are compensation coefficients for different periods of movable load movement.
[0159] In the process of setting the environmental protection objective function of this embodiment, CO2 emissions are used as the environmental protection index of the virtual power plant, where the carbon emissions of the virtual power plant mainly come from the main grid thermal power generation. The environmental protection cost generated during the operation of the virtual power plant is:
[0160]
[0161] Among them, C mp is the environmental cost; e is the carbon emission cost coefficient; E is the marginal emission factor.
[0162] In this embodiment, the constraints include, but are not limited to, system operation constraints, battery operation constraints, and user satisfaction constraints. System operation constraints ensure that power balance and grid interaction power are within a safe range. In this embodiment, battery operation constraints include, but are not limited to, charge and discharge power limits, number limits, and state of charge limits. In this embodiment, user satisfaction constraints take into account the impact of load scheduling on user electricity usage habits, measured by comfort and economy.
[0163] In this embodiment, the constraints of the optimization scheduling model include, but are not limited to: system operation constraints, battery operation constraints, and user satisfaction constraints.
[0164] In this embodiment, the system operation constraints include, but are not limited to: power balance equation constraints, grid interaction power constraints, and dispatch load power constraints.
[0165] In this embodiment, the power balance equation constraint is as follows:
[0166] P wt +P pv +P sb +P grid -P L =0
[0167] In the grid interaction power constraint of this embodiment, in the grid-connected operation state, the control output of the grid interaction power by the day-ahead load dispatch optimization should meet the line capacity limit and the main grid's restriction on the interaction power, that is:
[0168] -P gs,max ≤P grid ≤P gb,max
[0169] Among them, P gs,max P is the upper limit of the power sold by the virtual power plant to the main grid, expressed as a positive number; gb,max It is the upper limit of the power that the power grid can purchase from the main grid, expressed as a positive number.
[0170] In the scheduling load power constraint of this embodiment, in order to meet the basic power demand of users, the scheduling power of the load is limited, and the expression is:
[0171]
[0172] Among them, P M_out,max , P D,max They are the upper limit of the power of the removed load and the upper limit of the power of the disconnectable load respectively.
[0173] In this embodiment, the battery operation constraints include, but are not limited to: battery charge and discharge constraints, battery state constraints, and battery initial and final energy state constraints.
[0174] In the battery charge and discharge constraints of this embodiment, the battery charge and discharge power constraint condition is as follows:
[0175]
[0176] Among them, P sout,max is the maximum discharge power; P sin,max is the maximum charging power.
[0177] In this embodiment, the charge and discharge times of the battery are as follows:
[0178]
[0179] Where: R sin,t is the battery status during period t, 1 indicates that the battery is in a charging state, and 0 indicates that the battery is in a neither charging nor discharging state:
[0180] R sout,t R sin,t =0
[0181] In this embodiment, the battery state constraint is as follows:
[0182] S min t max
[0183] Among them, S max , S min They are the upper and lower limits of the state of charge (SOC) of the energy storage system (battery).
[0184] In this embodiment, the initial and final energy state constraints of the battery are as shown in the formula.
[0185] S0=S 24
[0186] Where S0 is the initial state of charge of the energy storage system (battery); S 24 It is the final state of charge value of the energy storage system (battery).
[0187] In the user satisfaction constraint of this embodiment, when the load responds, the dispatch of the movable load will directly affect the user's electricity usage habits, and the impact of load dispatch on the user can be measured by user satisfaction. User satisfaction can be divided into two parts: one is user comfort, and the other is user economy. In this embodiment, the user's electricity satisfaction is calculated based on the user's electricity comfort and economy, and the calculation formula is as follows:
[0188] M=U com U eco
[0189] Among them, M is user satisfaction; U com After optimizing for user load; U eco The economy of the user's electricity load.
[0190] In this embodiment, the user's comfort is calculated based on the difference between the user's actual power consumption curve and the maximum comfort curve. The calculation formula is as follows:
[0191]
[0192] In this embodiment, the economic degree of the user is calculated according to the difference in the cost of the user's electricity before and after load optimization, and the calculation formula is as follows:
[0193]
[0194] Among them, C bef is the user's electricity cost before load dispatch; C aft is the user's electricity cost after load scheduling. eco >1, the user's electricity cost is reduced after load optimization, and its economy is better than before load response; on the contrary, if U eco <1, then its economy is not as good as before the response. eco The larger the value, the better the economic benefits for users and the greater the user satisfaction value.
[0195] S3, applying the improved cuckoo algorithm to solve the model;
[0196] In this embodiment, an improved cuckoo search algorithm (CS) is used to solve the optimization scheduling problem of the virtual power plant. The algorithm simulates the parasitic behavior and Levy flight mode of the cuckoo, updates the position of the bird's nest (solution) in a random walk, and searches for the optimal solution. The Cuckoo search algorithm (CS) algorithm is inspired by the parasitic breeding habits of cuckoos in nature. It is a swarm intelligent search technology that takes into account the parasitic nature of cuckoo nests and the Levy flights mode. It changes the position of the bird's nest by random walks and has a strong global search capability.
[0197] In the CS algorithm of this embodiment that considers dynamically changing step sizes, the cuckoo search algorithm includes three basic assumptions:
[0198] First assumption: Assume that the cuckoo lays only one egg each time and randomly selects the nest location to incubate the egg;
[0199] Second assumption: In a cuckoo flock searching for a new nest, the bird that occupies the best position is able to lay the next generation of eggs in that nest;
[0200] The third assumption: The number of available bird nests is a constant. Now assume that the probability of a host bird finding cuckoo eggs is Pa, Pa∈[0,1].
[0201] In this embodiment, if cuckoo parasitism is discovered, the host bird will build a new nest nearby.
[0202] In this embodiment, the global search is performed by Levy flight, and the Levy flight position update formula is as follows:
[0203]
[0204] in, is the position of the jth (j=1, 2, 3, ..., n) bird's nest in the kth generation; is point-to-point multiplication; α is the step size control quantity, used to control the step size:
[0205]
[0206] (0 < s0 < s, where s0 is the minimum step size)
[0207] Among them,
[0208] Γ(λ) = ∫0 ∞ t λ-1 e -t dt
[0209] In this embodiment, the random numbers output by Levy flight are composed of a random number sequence generated by a normal distribution, and the step size is calculated through the Mantegna algorithm, as shown in the following formula:
[0210]
[0211] Among them, μ and ν follow a normal distribution:
[0212] and
[0213] and σ ν = 1
[0214] In this embodiment, the local search is performed through local random walks, and the position update formula is as shown in the following formula:
[0215]
[0216] Among them, and are the positions of two randomly selected nests in the current generation of nests; H(μ) is the Heaviside function; P a is the maximum discovery probability; s is the moving step size; ε is a random number uniformly distributed.
[0217] In this embodiment, the step size of the algorithm determines the search space of the algorithm for solutions. Generally, the step size of the cuckoo search algorithm is usually taken as α = 1. The fixed step size setting restricts the search for solutions to a certain extent. At the initial stage of the algorithm's operation, when the step size takes a larger value, the search space can be expanded to achieve better global search; when the algorithm runs in a smaller space, shortening the step size can be more conducive to local search and improve the search accuracy; therefore, the search efficiency of the algorithm can be enhanced by using a dynamically changing step size:
[0218]
[0219] Among them, a, b, c are step size control quantities, and a>b>c; k is the current iteration number; N is the total iteration number.
[0220] S4. Design and apply the CS algorithm with dynamically changing step size;
[0221] In the algorithm of this embodiment, a dynamically changing step size is introduced to enhance the search efficiency. The step size varies according to the number of iterations, with a larger step size in the early stage to expand the search space and a smaller step size in the later stage to improve the search accuracy.
[0222] S5, running the CS algorithm, and obtaining the economic optimization scheduling plan of the virtual power plant by finding the best-located bird's nest;
[0223] In the optimization scheduling problem of the virtual power plant in this embodiment, the original load curve and wind turbine and photovoltaic output of each time period have been clarified, but the charging and discharging status of the energy storage system (battery) and the movable load scheduling in each time period are uncertain. Therefore, in the optimization scheduling problem of the virtual power plant, the grid optimization planning can be carried out by planning the load response and battery action status in 24 time periods. When the cuckoo search algorithm is used to solve this problem, a bird's nest represents a group of battery charging and discharging status and load response status in 24 time periods. The quality of the bird's nest location represents the quality of the economic performance of the virtual power plant. The economic optimization scheduling plan of the virtual power plant is obtained by finding the best-positioned bird's nest.
[0224] like Figure 3 As shown, in this embodiment, the specific steps of running the CS algorithm include:
[0225] S51, input the parameters required by the algorithm and randomly initialize the bird's nest;
[0226] In the initialization process of this embodiment, the objective function is determined, the positions of the flocks of birds are initialized, and the algorithm parameters are set;
[0227] In this embodiment, the objective function f(x) is determined, the bird flock is initialized, and the initial positions of n bird nests are randomly generated.
[0228] x j (j=1,…,n). Set algorithm parameters: number of bird nests n, dimension n d , discovery probability P a , maximum number of iterations N and other parameters;
[0229] S52, calculating the objective function of the initialization bird's nest position, finding the optimal value and recording it;
[0230] S53, Levy flight to obtain the new nest location;
[0231] In the process of updating the position in this embodiment, the position of the bird's nest is updated according to the Levy flight and the local random walk, the advantages and disadvantages are calculated and the better position is retained;
[0232] In this embodiment, according to the following formula:
[0233]
[0234] Update the location of the current generation of bird nests, calculate the pros and cons of the location of the bird nest, compare it with the location of the previous generation of bird nests, and retain the bird nests with better locations;
[0235] S54, discarding some bird nests according to the discovery probability and looking for new bird nests nearby;
[0236] In this embodiment, in the case where parasitism is discovered, the nest position is randomly changed with a specific probability of discovery;
[0237] S55, calculating the objective function of the new bird's nest position and comparing it with the last optimal value to obtain a new optimal value;
[0238] In this embodiment, the random number r is used as the probability that the owner of the bird nest will find the foreign bird eggs, and it is compared with the probability that the bird will be eliminated.
[0239] P a
[0240] For comparison.
[0241] r>P a Then randomly change the nest positions to obtain a new set of nest positions, and keep the better nest positions;
[0242] S56, judging whether the iteration end condition is met;
[0243] In this embodiment, an iterative judgment is performed, and if the maximum number of iterations is reached, the optimal solution is output, otherwise step S52 is repeated for iteration.
[0244] In this embodiment, it is determined whether the current number of iterations meets the set maximum number of iterations. If it does, the search process ends and the global optimal value is output.
[0245] f min and the global optimal position; otherwise, repeat the aforementioned step S52 to perform iterative optimization.
[0246] The present application provides a virtual power plant optimization scheduling method based on dual-mode economic model predictive control. This method constructs a multi-objective optimization scheduling model that comprehensively considers economic benefits, environmental benefits, system stability and user satisfaction by introducing an improved cuckoo search algorithm and a two-way raiding multi-objective gray wolf algorithm, which significantly improves the global search capability and convergence speed of the algorithm and enhances the adaptability of the virtual power plant to the volatility of new energy. In addition, the present application also proposes an intelligent energy storage system scheduling strategy that can be optimized and adjusted according to the actual needs and operating status of the power grid to improve the utilization efficiency of the energy storage system and the overall performance of the virtual power plant. Through the adaptive step size strategy and dual-mode economic model predictive control, the present application not only improves the operating economy of the virtual power plant while ensuring good dynamic tracking performance of the set point, but also solves the optimization algorithm limitations, insufficient demand-side management, and the singleness of the energy storage system scheduling strategy in the prior art. The problem provides strong technical support for the construction and sustainable development of smart grids.
[0247] In summary, the present invention constructs a multi-objective optimization scheduling model that comprehensively considers economic benefits, environmental benefits, system stability and user satisfaction. Through the improved cuckoo algorithm, the optimal scheduling of the source-load-storage power of the virtual power plant under multiple constraints is achieved.
[0248] The present invention proposes a more flexible and intelligent energy storage system scheduling strategy, which can be optimized and adjusted according to the actual needs and operating status of the power grid, improving the utilization efficiency of the energy storage system and the overall performance of the virtual power plant. The present invention not only solves some key problems in the optimization and scheduling of virtual power plants, but also improves the solution efficiency and solution quality by improving the algorithm. This research result has important theoretical and practical significance for promoting the development and application of virtual power plant technology.
[0249] The present invention introduces a demand-side response mechanism and utilizes the scheduling of movable loads and disconnectable loads to optimize power demand in different time periods, reduce the operating costs of virtual power plants, and improve the stability and reliability of the system.
[0250] The present invention constructs a virtual power plant active power dispatching model that comprehensively considers system operation constraints, battery operation constraints and user satisfaction. The improved cuckoo algorithm of this study can effectively deal with the source-load-storage coordination problem in the virtual power plant.
[0251] In the model of the present invention, we not only consider the uncertainty of wind and photovoltaic power generation, but also the dynamic changes of load and the operating limitations of energy storage equipment. Through application and verification in multiple typical scenarios, the results show that this method can effectively reduce the operating cost of virtual power plants while improving the environmental performance of the system.
[0252] The present invention constructs a virtual power plant power dispatch model that includes wind and solar power generation, energy storage units and loads, comprehensively considers system operation constraints, battery operation constraints and user satisfaction, and takes the optimization of the economic and environmental performance of the virtual power plant as the objective function, thereby improving the adaptability of the virtual power plant to the volatility of renewable energy.
[0253] The present invention improves the traditional cuckoo algorithm and introduces an adaptive step size strategy, so that the algorithm can dynamically adjust the search step size according to the current solution during the search process, thereby achieving a better balance between global search and local search.
[0254] The improved cuckoo algorithm of the present invention not only improves the quality of the solution when solving the virtual power plant optimization scheduling problem, but also outperforms other algorithms in terms of computational efficiency. The improved cuckoo algorithm is an effective tool for virtual power plant optimization scheduling, which can provide strong support for the economic and environmental optimization of virtual power plants.
[0255] The present invention improves the global search capability and convergence speed of the algorithm through adaptive step size adjustment and Levy flight strategy, thereby more effectively solving the optimization scheduling problem of the virtual power plant.
[0256] The present invention not only improves the economic operation efficiency of virtual power plants, but also enhances their adaptability to the volatility of renewable energy, realizes the optimization of demand-side management, improves the scheduling efficiency of energy storage systems, and finds the best balance point among multiple objectives, providing strong technical support for the sustainable development of virtual power plants and the construction of smart grids.
[0257] The present invention solves the technical problems existing in the prior art, namely, the optimization algorithm is limited, lacks adaptability to the volatility of renewable energy, inadequate demand-side management, the singleness of the energy storage system scheduling strategy, and lacks a scheduling model that comprehensively considers multiple objectives.
[0258] The above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit the same. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that the technical solutions described in the aforementioned embodiments may still be modified, or some of the technical features may be replaced by equivalents. However, these modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the embodiments of the present invention.
Claims
1. A virtual power plant source-load-storage power optimization scheduling method, characterized in that: The method comprises: S1. Construct a virtual power plant operation model, including: a distributed power generation unit model, an energy storage unit model and a load unit model; in the virtual battery operation model, divide the user load in the power supply model into fixed load, movable load and disconnectable load for scheduling optimization, wherein the distributed power generation unit model includes: a wind power generation model and a photovoltaic power generation model; S2. Construct a virtual battery optimization scheduling model, including an economic objective function and an environmental objective function, and set constraints, including: system operation constraints, battery operation constraints, and user satisfaction constraints; the system operation constraints include: power balance equation constraints and grid interaction power constraints; wherein the battery operation constraints include: battery charging and discharging constraints, battery state constraints, and battery initial and final energy state constraints; S3. Adjust and obtain an applicable cuckoo search algorithm CS to solve the virtual power plant operation model and the virtual battery optimization scheduling model; wherein, using the applicable cuckoo search algorithm CS, the cuckoo parasitic behavior is simulated and Levy flight is performed, global search and local search are performed, and the bird's nest solution position is updated to obtain the optimal solution of the model; S4. Run the applicable cuckoo search algorithm CS to obtain an economic optimization scheduling plan for the virtual power plant based on the optimal solution of the model.
2. The virtual power plant source-load-storage power optimization scheduling method according to claim 1 is characterized in that: The S1 includes: S11. Construct the wind power generation model according to the following wind power output power determination logic: Where P wt is the fan output power, ρ is the air density, R is the rotation radius of the fan blade, ν is the maximum linear velocity on the fan blade, C p is the wind power conversion coefficient; S12. Construct the photovoltaic power generation model according to the following photovoltaic output power determination logic: Where P pv , L, T pv are the output power, light intensity and temperature of photovoltaic power generation under the target environment, P STC , L STC , T STC are the output power, light intensity and temperature of photovoltaic power generation under standard test conditions, k T is the power temperature coefficient; S13, constructing the energy storage unit model using the following state of charge SOC value determination logic: In the formula, S t is the SOC value of the battery during period t, P sb,t is the charging and discharging power of the energy storage system in the tth period, E b is the rated capacity of the battery, and Δt is the time period. S14. Construct the load unit model using the following total load determination logic: P L (t)=P B (t)+P M (t)+P D (t) Where P L (t) is the total load power in the virtual power plant during period t; P B (t), P M (t) and P B (t) are the fixed load power, movable load power and disconnectable load power in period t respectively. S15. Model the movable load using the following logic: Where P M_in,i (t) is the transferable load power from period i to period t; P M_out (t) is the load power that can be removed during period t.
3. The virtual power plant source-load-storage power optimization scheduling method according to claim 1 is characterized in that: The S2 includes: S21. Using the following virtual power plant economic cost logic, the economic objective function is set: minC mg =C wt +C pv +C sb +C grid +C D +C M In the formula, C mg is the economic cost of the microgrid, C wt is the cost of wind power generation, C pv is the photovoltaic power generation cost, C sb is the battery operating cost, C grid is the cost of exchanging electricity between the virtual power plant and the main grid, C D Compensation for users’ power interruption, C M Dispatch and compensate for users' electricity load; S22. Using the following virtual power plant environmental cost determination logic, set the environmental protection objective function: In the formula, C mp is the environmental cost, e is the carbon emission cost coefficient, and E is the marginal emission factor. S23. Determine the power balance equation constraint using the following logic: P wt +P pv +P sb +P grid -P L =0 S24. Determine the grid interaction power constraint using the following logic: -P gs,max ≤P grid ≤P gb,max Where P gs,max P is the upper limit of the power sold by the virtual power plant to the main grid, expressed as a positive number; gb,max The upper limit of the power that the power grid can purchase from the main grid, expressed as a positive number; S25. Determine the dispatch load power constraint using the following logic to limit the load dispatch power: Where P M_out,max , P D,max They are the upper limit of the power of the removed load and the upper limit of the power of the disconnectable load respectively; S26. Determine the battery charge and discharge constraints using the following logic: Where P sout,max is the maximum discharge power, P sin,max is the maximum charging power; S27. Determine the battery state constraint using the following logic: S min <S t <S max In the formula, S max , S min They are the upper and lower limits of the state of charge (SOC) of the energy storage system respectively. S28. Determine the user satisfaction constraint using the following logic: M=U com U eco In the formula, M is user satisfaction, U is com After optimizing for user load, U eco The economic degree of the user's electricity load; S29, obtaining and calculating the user comfort level based on the difference between the user's actual power consumption curve and the maximum comfort level curve using the following logic: S210, according to the difference in the cost of electricity used by the user before and after load optimization, the user's economic degree is calculated using the following logic: In the formula, C bef is the user's electricity cost before load dispatch, C aft The electricity cost of users after load scheduling.
4. The virtual power plant source-load-storage power optimization scheduling method according to claim 3 is characterized in that: In S21, the following logic is used to determine the economic cost of the microgrid, the wind power generation cost, the photovoltaic power generation cost, the battery operation cost, the cost of exchanging electric energy between the virtual power plant and the main grid, the user's power interruption compensation, and the user's power load scheduling compensation: In the formula, α wt is the unit wind power generation cost; α pv is the unit photovoltaic power generation cost; R sout,t is the battery status during period t; β sb is the unit discharge cost of the battery; g B P is the unit electricity purchase cost of the microgrid during period t; grid is the interaction power between microgrid and main grid; D is the interruptible load compensation coefficient; a t,i Unit dispatch compensation for load shifting from period i to period t; The unit dispatch compensation a of the load moving from period i to period t is determined by the following logic: t,i : Where a1, a2, a3, and a4 are the compensation coefficients of the movable load at different moving periods.
5. The virtual power plant source-load-storage power optimization scheduling method according to claim 1 is characterized in that: In step S26, the following logic is used to determine the battery charge and discharge times: In the formula, R sin,t is the battery status during period t.
6. The virtual power plant source-load-storage power optimization scheduling method according to claim 1 is characterized in that: The S3 includes: S31, performing the global search through the Levy flight to update the Levy flight position using the following logic: In the formula, is the position of the jth (j=1,2,3,…,n) bird’s nest in the kth generation, is point-to-point multiplication, α is the step size control value, which is used to control the step size; S32, performing the local search by local random walk: In the formula, and are the positions of two random nests in the contemporary nest, H(μ) is the Heaviside function, P a is the maximum discovery probability, s is the moving step length, and ε is a uniformly distributed random number. S33, using the following logic, using the dynamically changing step size enhancement algorithm to adjust the step size: Where a, b, c are step size control quantities, and a>b>c; k is the current iteration number; N is the total iteration number.
7. The method for optimizing the dispatching of source, load and storage power of a virtual power plant according to claim 6, characterized in that: In step S31, the step size is determined using the following logic:
8. The virtual power plant source-load-storage power optimization scheduling method according to claim 1 is characterized in that: In step S31, the step length is calculated by the Mantegna algorithm:
9. The virtual power plant source-load-storage power optimization scheduling method according to claim 1 is characterized in that: In step S31, μ and ν obey the following normal distribution: and Japanese v =1.
10. The virtual power plant source-load-storage power optimization dispatching system is characterized by: The system comprises: A power plant operation modeling module is used to construct a virtual power plant operation model, including: a distributed power generation unit model, an energy storage unit model and a load unit model; in the virtual battery operation model, the user load in the power supply model is divided into fixed load, movable load and disconnectable load for scheduling optimization, wherein the distributed power generation unit model includes: a wind power generation model and a photovoltaic power generation model; The battery scheduling modeling module is used to construct a virtual battery optimization scheduling model, including an economic objective function, an environmental objective function, and setting constraints, including: system operation constraints, battery operation constraints, and user satisfaction constraints; the system operation constraints include: power balance equation constraints and grid interaction power constraints; wherein the battery operation constraints include: battery charging and discharging constraints, battery state constraints, and battery initial and final energy state constraints; A model solving module, used for adjusting and obtaining an applicable cuckoo search algorithm CS, so as to solve the model of the virtual power plant operation model and the virtual battery optimization scheduling model; wherein, the applicable cuckoo search algorithm CS is used to simulate the cuckoo parasitic behavior and perform Levy flight, perform global search and local search, and update the bird's nest solution position to obtain the optimal solution of the model, and the model solving module is connected with the power plant operation modeling module and the battery scheduling modeling module; The scheduling scheme determination module is used to run the applicable cuckoo search algorithm CS and obtain the economic optimization scheduling scheme of the virtual power plant according to the optimal solution of the model. The scheduling scheme determination module is connected to the model solving module.
Citation Information
Patent Citations
Virtual power plant optimal scheduling method, device and equipment based on big data, and medium
CN114037191A
Virtual power plant optimal scheduling method based on dual-mode economic model predictive control
CN117151357A
Virtual power plant optimization scheduling method considering electric vehicle demand response and application thereof
CN117543581A
Load side virtual power plant scheduling method based on improved swarm intelligence algorithm
CN117595239A
Virtual power plant distributed photovoltaic energy storage system configuration optimization method and device
CN118070988A
Cited By
Dynamic correction type source load model data dual-drive power grid coordination control method and system
CN119561158A
Electro-hydrogen fusion time sequence production simulation method
CN119783406A
Intelligent scheduling method and system for cross-regional load migration of virtual power plant
CN120165448A
Intelligent scheduling method and system for cross-regional load migration of virtual power plants
CN120165448B
Electrified road optimization scheduling method based on virtual power plant
CN120767824A
