Energy storage type virtual power plant energy distribution method considering green power transaction and new energy consumption

Abstract

The invention discloses an energy storage type virtual power plant energy distribution method considering green power transaction and new energy consumption, and the method comprises the steps: building an uncertainty model of each random variable under a green power transaction condition through a scene random differential method through generating a classification scene data set of frequency modulation demands and wind and light output original data under a green power transaction environment; generating a frequency modulation demand and a wind and light output scene with time sequence correlation; with the optimal benefit of the energy storage type virtual power plant as a target function, establishing an energy management two-stage stochastic optimization model of the energy storage type virtual power plant participating in the electric energy-frequency modulation combined market, and setting energy management model constraint conditions of the energy storage type virtual power plant participating in the electric energy-frequency modulation combined market; and solving the energy management model to obtain an optimization scheme that the energy storage type virtual power plant participates in power market profit. According to the method, the specific flexible adjustment characteristic of the energy storage type virtual power plant in a novel power system can be exerted, the scheduling plan performability is improved, the income mode of the virtual power plant is enriched, and the income level of the independent virtual power plant participating in the green electricity transaction market is effectively improved.

Description

Technical Field

[0001] The present invention relates to a technology in the field of power resource allocation, specifically a method for allocating energy of a storage-type virtual power plant taking into account green electricity trading and new energy consumption. Background Art

[0002] A high proportion of renewable energy power generation and grid-connected systems will become an inevitable trend in the development of new power systems in the future. Energy storage virtual power plants that include high proportions of energy storage facilities have the characteristics of high operational flexibility, fast response time, and high ramp rate. They can support the operation of new power systems with high proportions of renewable energy power generation by adjusting their own charging or discharging behaviors. Their role in maintaining the balance of supply and demand in power systems has been widely recognized. With the increasing demand for energy storage systems on the source, grid, and load sides and the rise of the concept of sharing economy, the business model of energy storage virtual power plants that provide services to multiple entities has gradually attracted attention. Summary of the invention

[0003] In view of the problems of energy management of existing energy storage type virtual power plants, limited profit levels from participating in green electricity trading and spot markets, and the impact of multiple uncertainties on their promotion of new energy consumption, the present invention proposes an energy allocation method for energy storage type virtual power plants that takes green electricity trading and new energy consumption into consideration. The method can give full play to the flexible adjustment characteristics unique to energy storage type virtual power plants in new power systems, improve the executability of scheduling plans, enrich the profit model of virtual power plants, and effectively improve the profit level of independent virtual power plants participating in the green electricity trading market.

[0004] The present invention is achieved through the following technical solutions:

[0005] The present invention relates to an energy allocation method for an energy storage type virtual power plant taking into account green electricity trading and new energy consumption. The method generates a classified scenario data set of original data of frequency regulation demand and wind and solar power output under a green electricity trading environment, and constructs an uncertainty model of each random variable under green electricity trading conditions through a scenario random differential method, thereby generating frequency regulation demand and wind and solar power output scenarios with time series correlation. The method then takes the optimal profit of the energy storage type virtual power plant itself as the objective function, establishes a two-stage random optimization model of energy management of the energy storage type virtual power plant participating in an electric energy-frequency regulation joint market, and sets constraints on the energy management model of the energy storage type virtual power plant participating in the electric energy-frequency regulation joint market. The method solves the energy management model to obtain an optimization plan for the energy storage type virtual power plant to make profits by participating in the electricity market.

[0006] The present invention relates to a system for implementing the above method, comprising: a data management unit, an uncertainty characterization unit and an energy management unit, wherein: the data management unit collects and processes frequency modulation requirements and wind and solar power output data under green electricity trading conditions to obtain a typical scenario in which a virtual power plant is located; the uncertainty characterization unit uses a stochastic differential method to characterize multiple uncertainties for direct identification and processing by the virtual power plant energy management unit; the energy management unit optimizes the energy release of each device within the energy storage type virtual power plant based on all information. Technical Effects

[0007] The present invention uses a multiple uncertainty scenario generation and analysis method based on stochastic differential equations. On the basis of considering the uncertainty of traditional wind and solar power and the market electricity price, it considers the frequency regulation uncertainty caused by the imbalance of wind and solar power in green power trading, and uses the scenario generation method of the stochastic differential method to effectively characterize the multiple uncertainty scenarios faced by energy storage virtual power plants, which has helped virtual power plants to manage energy reasonably and efficiently. Compared with the prior art, the present invention uses a multiple uncertainty scenario generation and analysis method based on stochastic differential equations, which can make the generated scenario data more concentrated in autocorrelation (Wasserstein distance distribution), with the smallest value, more similar to the actual historical data distribution, and more in line with the actual data. BRIEF DESCRIPTION OF THE DRAWINGS

[0008] Figure 1 It is a flow chart of the present invention;

[0009] Figure 2 A schematic diagram of a random variable scenario set generated by using a scenario random differentiation method in an embodiment;

[0010] Figure 3 Comparison chart of the autocorrelation characteristics between the signal generated by the scenario random differentiation method and the scenario generation method predicted by the chance LSTM and the actual signal;

[0011] Figure 4 The results of using the energy management method of the present invention for energy storage type virtual power plants to participate in various markets;

[0012] Figure 5 The results of participating in various markets for energy storage virtual power plants using energy management methods that do not provide balancing services for new energy stations;

[0013] Figure 6 Results of participating in various markets for energy storage virtual power plants using an energy management approach that does not consider uncertainty. DETAILED DESCRIPTION

[0014] like Figure 1 As shown, this embodiment relates to an energy allocation method for an energy storage type virtual power plant considering green electricity trading and new energy consumption, including the following steps:

[0015] Step 1: Generate a classification scenario data set considering the frequency regulation demand under green power trading and the original data of wind and solar power output. Specifically, the K-Means clustering analysis method is used to classify the original data of frequency regulation demand under green power trading and wind and solar power output to obtain the classified original data set, in which the frequency regulation demand is represented by the market frequency regulation signal data, such as Figure 2 shown.

[0016] Step 2: Based on the classification scenario data set and the principle of stochastic differential equations, a scenario stochastic differential method is proposed to construct the uncertainty model of each random variable in the green electricity trading environment, including:

[0017] ① For frequency modulation demand, the market frequency modulation signal data is used to characterize it, and the stochastic differential equation model dξ of the frequency modulation signal is constructed using the parameter estimation method based on the Fokker-Planck equation t =μ(t,ξ t )dt+σ(t,ξ t )dW t , where: μ(ξ t )=-λξ t , ξ t is the random variable to be described, W t is the Wiener motion, μ(t,ξ t ) is the drift coefficient, which represents the random variable ξ t The trend of moving toward the expected value, σ(t,ξ t ) is the diffusion coefficient W t The randomness of ξ t impact.

[0018] ② For wind and solar power output, the prediction error sequence {e0(t),e0(t),…,e N (t)} is characterized and the maximum likelihood estimation method is used to obtain the stochastic differential equation model of wind and solar power output: de(t)=θ1e(t)dt+θ2dW t dξ t =μ(t,ξ t )dt+σ(t,ξ t )dW t , for the forecast error sequence {e0(t),e0(t),…,e N (t)}, using the maximum likelihood estimation method, the maximum likelihood function of the parameter θ is The maximum likelihood function takes partial derivatives of the two parameters θ1 and θ2 as 0, then the parameters

[0019] Step 3: Generate frequency regulation demands and wind and solar output scenarios with time correlation based on the uncertainty model, take the optimal profit of the energy storage virtual power plant as the objective function, use energy storage to provide balancing services for wind power and photovoltaic power stations, promote the consumption of new energy, and establish a two-stage stochastic optimization model for energy management of energy storage virtual power plants participating in the electricity-frequency regulation joint market.

[0020] The frequency modulation demand and wind and solar power output scenarios with time-series correlation refer to: combining the scenario analysis method with the stochastic differential equation, considering the time-series correlation between the influencing factors in the scenario, and obtaining the uncertainty scenarios of the three random variables of frequency modulation signal, wind power output and photovoltaic output.

[0021] The use of energy storage to provide balancing services for wind power and photovoltaic power stations and promote the consumption of new energy means that energy storage virtual power plants use their internal high-proportion energy storage facilities to charge or discharge the actual output and predicted output deviation of new energy such as wind power and photovoltaic power, so as to reduce the deviation penalty costs caused by the fluctuation of new energy output. At the same time, this behavior can reduce the imbalance of green electricity such as wind power and photovoltaic power participating in market transactions and improve the power grid's ability to consume new energy power generation.

[0022] The two-stage stochastic optimization model of energy management is specifically:

[0023] The overall goal of energy management is to maximize the benefits of the virtual power plant by allocating energy discharge processes in different time periods: Among them: Electric energy income in the day-ahead stage is the energy market price in the day-ahead market, P t en,c , P t en,d They are the day-ahead charging and discharging power reported by the energy storage virtual power plant in the energy market; the frequency regulation income of the energy storage virtual power plant are the frequency regulation capacity and mileage price in the day-ahead market, P t reg Day-ahead power reporting for energy storage virtual power plants in the frequency regulation market; balancing service revenue is the agreed day-ahead rental price; the electricity revenue of the virtual power plant in real time is the real-time electricity market price, They are respectively the positive and negative imbalance of charging and discharging power reported by the energy storage virtual power plant in the electric energy market at time s and τ in the scenario; the benefits obtained by the energy storage virtual power plant in frequency regulation They are the frequency regulation market capacity price and the corresponding mileage price in the real-time stage. is the market declaration deviation penalty coefficient, m is the average mileage call rate, are the positive and negative imbalances of the frequency regulation capacity declared by the energy storage virtual power plant at time τ in scenario s. Balancing service income is the real-time electricity market price, are respectively the positive and negative imbalance of charging and discharging power reported by the energy storage virtual power plant in the electric energy market at time s and τ in the scenario; the initial investment discount cost of the energy storage virtual power plant Operating life loss cost c cap is the unit price of capacity cost at the initial investment of energy storage equipment, is the rated capacity of internal energy storage, r is the average investment discount rate of energy storage projects, T life For the overall life cycle of energy storage, are the total charging and discharging power of the energy storage at the scene s and time τ, respectively, es is the energy storage operation loss coefficient, It represents the charging and discharging energy generated by the energy storage virtual power plant in response to the frequency modulation signal during the dispatch period.

[0024] Step 4: Construct the energy management model constraints for the energy storage virtual power plant to participate in the electricity energy-frequency regulation joint market, including:

[0025] 4.1) Constructing the operating power constraints of energy storage virtual power plants Where: P max is the maximum charging and discharging power of the energy storage virtual power plant; P t c , P t d They are the charging and discharging power of the energy storage virtual power plant respectively.

[0026] 4.2) Constructing energy storage virtual power plant to balance service power constraints in: is a 0-1 variable, which is used to avoid the simultaneous charging and discharging of energy storage in the energy storage virtual power plant at a certain moment; P t pv,rt , P t pv,da , P t wt,rt , P t wt,da Provide real-time and day-ahead forecasts for photovoltaic power plants and wind power companies.

[0027] 4.3) Construct energy storage SOC constraints; SOC t =SOC t-1 +ΔSOC t,t∈{1,...,T-1}, SOC min ≤SOC t ≤SOC max ,SOC min ≤S T =S0≤SOC max , where: SOC max , SOC min They are respectively the maximum and minimum charge states of the energy storage equipment in the energy storage virtual power plant.

[0028] 4.4) Constructing frequency regulation state constraints of energy storage virtual power plants in: It is the charging and discharging energy generated by the energy storage virtual power plant during the dispatching period following the frequency modulation signal; is the cumulative effect value of the frequency modulation signal, Δt is the length of the scheduling period, and Δt reg is the frequency modulation signal period, K is the number of frequency modulation signals in the scheduling period Δt; SOC lb , SOC ub They are the upper and lower limits of SOC that energy storage virtual power plants are allowed to participate in the frequency regulation market.

[0029] 4.5) There are association constraints between the two-stage decision variables:

[0030]

[0031] Step 5: Solve the energy management model of the energy storage virtual power plant to obtain the optimization parameters for the energy storage virtual power plant to participate in the electricity market and make profits. Specifically, Matlab is selected to program the energy management model, and the commercial solver Gurobi is used to solve it. The optimization parameters obtained by calculation include: Where: obj is the overall profit value of the energy storage virtual power plant.

[0032] After specific experiments, using the day-ahead and real-time market price data and frequency modulation signal data from the PJM market on a certain trading day, and the actual historical power generation data of wind power and photovoltaic power generation in a certain area of ​​the Elia power grid, by setting up a storage-type virtual power plant, the energy management method of the present invention, the management method that does not consider providing balancing services for new energy stations, and the management method that does not consider the impact of uncertainty are three situations, such as Figure 3 As shown, compared with other methods, the present invention has a higher degree of concentration in the Wasserstein distance distribution, takes the smallest value, is more similar to the actual historical data distribution, and is more in line with the actual data, as shown in Table 1.

[0033] Table 1 Comparison of Wasserstein distance of scene generation methods

[0034] As shown in Table 2 and Figure 4-6 As shown, the energy storage virtual power plant ensures the scheduling requirements before they are actually feasible through the energy management of this method, while obtaining greater benefits.

[0035] Table 2 Comparison of energy management effects of energy storage virtual power plants

[0036] In summary, this method provides a new energy management method for energy storage virtual power plants, reduces the occurrence of practically unfeasible situations in energy management, and improves the safety and reliability of energy management. At the same time, while giving priority to safety, this method further improves the economic benefits of energy storage virtual power plants and stimulates enthusiasm for their investment and construction.

[0037] The above-mentioned specific implementation can be partially adjusted in different ways by those skilled in the art without departing from the principle and purpose of the present invention. The protection scope of the present invention shall be based on the claims and shall not be limited by the above-mentioned specific implementation. Each implementation scheme within its scope shall be subject to the constraints of the present invention.

Claims

1. A method for allocating energy to a virtual power plant with energy storage that takes into account green electricity trading and new energy consumption, characterized in that: By generating a classified scenario data set of original data on frequency regulation demand and wind and solar power output under the green electricity trading environment, the uncertainty model of each random variable under the green electricity trading conditions is constructed through the scenario stochastic differentiation method, and the frequency regulation demand and wind and solar power output scenarios with time series correlation are generated; then, taking the optimal profit of the energy storage virtual power plant itself as the objective function, a two-stage stochastic optimization model of energy management of the energy storage virtual power plant participating in the electric energy-frequency regulation joint market is established, and the energy management model constraints of the energy storage virtual power plant participating in the electric energy-frequency regulation joint market are set; by solving the energy management model, the optimization plan for the energy storage virtual power plant to make profits by participating in the electricity market is obtained.

2. The energy allocation method of energy storage type virtual power plant considering green power trading and new energy consumption according to claim 1 is characterized in that: include: Step 1: Generation of a classified scenario data set considering the frequency regulation demand under green electricity trading and the original data of wind and solar power output, specifically: using the K-Means clustering analysis method to classify the original data of frequency regulation demand under green electricity trading and wind and solar power output, and obtaining a classified original data set, in which the frequency regulation demand is represented by the market frequency regulation signal data, as shown in Figure 2; Step 2: Based on the classification scenario data set and the principle of stochastic differential equations, a scenario stochastic differential method is proposed to construct the uncertainty model of each random variable in the green electricity trading environment, including: ① For frequency modulation demand, the market frequency modulation signal data is used to characterize it, and the stochastic differential equation model dξ of the frequency modulation signal is constructed using the parameter estimation method based on the Fokker-Planck equation t =μ(t,ξ t )dt+σ(t,ξ t )dW t , where: μ(ξ t )=-λξ t , ξ t is the random variable to be described, W t is the Wiener motion, μ(t,ξ t ) is the drift coefficient, which represents the random variable ξ t The trend of moving toward the expected value, σ(t,ξ t ) is the diffusion coefficient W t The randomness of ξ t The impact of ② For wind and solar power output, the prediction error sequence {e0(t),e0(t),…,e N (t)} is characterized and the maximum likelihood estimation method is used to obtain the stochastic differential equation model of wind and solar power output: de(t)=θ1e(t)dt+θ2dW t dξ t =μ(t,ξ t )dt+σ(t,ξ t )dW t , for the forecast error sequence {e0(t),e0(t),…,e N (t)}, using the maximum likelihood estimation method, the maximum likelihood function of the parameter θ is The maximum likelihood function takes partial derivatives of the two parameters θ1 and θ2 as 0, then the parameters Step 3: Generate frequency regulation demand and wind and solar output scenarios with time series correlation based on the uncertainty model, take the optimal benefit of the energy storage virtual power plant as the objective function, use energy storage to provide balancing services for wind power and photovoltaic power stations, promote the consumption of new energy, and establish a two-stage stochastic optimization model for energy management of energy storage virtual power plants participating in the electricity-frequency regulation joint market; Step 4: Construct the energy management model constraints for the energy storage virtual power plant to participate in the electricity energy-frequency regulation joint market, including: 4.1) Constructing the operating power constraints of energy storage virtual power plants Where: P max is the maximum charging and discharging power of the energy storage virtual power plant; P t c , P t d They are the charging and discharging power of the energy storage virtual power plant respectively; 4.2) Constructing energy storage virtual power plant to balance service power constraints in: is a 0-1 variable, which is used to avoid the simultaneous charging and discharging of energy storage in the energy storage virtual power plant at a certain moment; P t pv,rt , P t pv,da , P t wt,rt , P t wt,da Provide real-time and day-ahead forecasts for photovoltaic power plants and wind power companies; 4.3) Construct energy storage SOC constraints; SOC t =SOC t-1 +ΔSOC t ,t∈{1,...,T-1}, SOC min ≤SOC t ≤SOC max ,SOC min ≤S T =S0≤SOC max , where: SOC max , SOC min They are the maximum and minimum charge states of the energy storage equipment in the energy storage virtual power plant respectively; 4.4) Constructing frequency regulation state constraints of energy storage virtual power plants in: It is the charging and discharging energy generated by the energy storage virtual power plant during the dispatching period following the frequency modulation signal; is the cumulative effect value of the frequency modulation signal, Δt is the length of the scheduling period, and Δt reg is the frequency modulation signal period, K is the number of frequency modulation signals in the scheduling period Δt; SOC lb , SOC ub They are the upper and lower limits of SOC that energy storage virtual power plants are allowed to participate in the frequency regulation market; 4.5) There are association constraints between the two-stage decision variables: Step 5: Solve the energy management model of the energy storage virtual power plant to obtain the optimization parameters for the energy storage virtual power plant to participate in the electricity market and make profits. Specifically, Matlab is selected to program the energy management model, and the commercial solver Gurobi is used to solve it. The optimization parameters obtained by calculation include: Where: obj is the overall profit value of the energy storage virtual power plant.

3. The energy allocation method of energy storage type virtual power plant considering green electricity trading and new energy consumption according to claim 2 is characterized in that: The frequency modulation demand and wind and solar power output scenarios with time-series correlation refer to: combining the scenario analysis method with the stochastic differential equation, considering the time-series correlation between the influencing factors in the scenario, and obtaining the uncertainty scenarios of the three random variables of frequency modulation signal, wind power output and photovoltaic output.

4. The energy allocation method of energy storage type virtual power plant considering green electricity trading and new energy consumption according to claim 2 is characterized in that: The use of energy storage to provide balancing services for wind power and photovoltaic power stations and promote the consumption of new energy means that the energy storage virtual power plant uses the high-proportion energy storage facilities within it to charge or discharge the deviation between the actual output and the predicted output of new energy such as wind power and photovoltaics, so as to reduce the deviation penalty costs caused by the fluctuation of new energy output. At the same time, this behavior can reduce the imbalance of green electricity such as wind power and photovoltaics participating in market transactions and improve the power grid's ability to absorb new energy power generation.

5. The energy allocation method of energy storage type virtual power plant considering green electricity trading and new energy consumption according to claim 2 is characterized in that: The two-stage stochastic optimization model of energy management is specifically: The overall goal of energy management is to maximize the benefits of the virtual power plant by allocating energy discharge processes in different time periods: Among them: Electric energy income in the day-ahead stage is the energy market price in the day-ahead market, P t en,c , P t en,d They are the day-ahead charging and discharging power reported by the energy storage virtual power plant in the energy market; the frequency regulation income of the energy storage virtual power plant are the frequency regulation capacity and mileage price in the day-ahead market, P t reg Day-ahead power reporting for energy storage virtual power plants in the frequency regulation market; balancing service revenue is the agreed day-ahead rental price; the electricity revenue of the virtual power plant in real time is the real-time electricity market price, They are respectively the positive and negative imbalance of charging and discharging power reported by the energy storage virtual power plant in the electric energy market at time s and τ in the scenario; the benefits obtained by the energy storage virtual power plant in frequency regulation They are the frequency regulation market capacity price and the corresponding mileage price in the real-time stage. is the market declaration deviation penalty coefficient, m is the average mileage call rate, are the positive and negative imbalances of the frequency regulation capacity declared by the energy storage virtual power plant at time τ in scenario s; the balancing service income is the real-time electricity market price, are respectively the positive and negative imbalance of charging and discharging power reported by the energy storage virtual power plant in the electric energy market at time s and τ in the scenario; the initial investment discount cost of the energy storage virtual power plant Operating life loss cost c cap is the unit price of capacity cost at the initial investment of energy storage equipment, is the rated capacity of internal energy storage, r is the average investment discount rate of energy storage projects, T life For the overall life cycle of energy storage, are the total charging and discharging power of the energy storage at the scene s and time τ, respectively, es is the energy storage operation loss coefficient, It represents the charging and discharging energy generated by the energy storage virtual power plant in response to the frequency modulation signal during the dispatch period.

6. A system for implementing the energy allocation method of energy storage type virtual power plant considering green electricity trading and new energy consumption as described in any one of claims 1-5, characterized in that: include: Data management unit, uncertainty characterization unit and energy management unit, among which: the data management unit collects and processes the frequency regulation demand and wind and solar power output data under green electricity trading conditions to obtain the typical scenario of the virtual power plant; the uncertainty characterization unit uses the stochastic differential method to characterize multiple uncertainties for direct identification and processing by the virtual power plant energy management unit; the energy management unit optimizes the energy release of each device within the energy storage virtual power plant based on all information.

Images