Two-Layer Optimal Operation Method of Distribution Network Energy Storage Based on Stepwise Price Demand Response

Abstract

The present invention discloses a two-layer optimal operation method for distribution network energy storage based on stepped price demand response. The method comprises the following steps: S1. Considering the uncertainty of real-time electricity price fluctuations, an autoregressive moving average time series model and an interval optimization method are used to establish a dynamic real-time electricity price model; S2. By analyzing the relationship between electricity price and demand response, a stepped price-demand response model is established; S3. Combining the charge and discharge capacity and efficiency constraints of battery energy storage, a two-layer optimization model of the stepped price-based demand response mechanism and regional energy storage operation is established. The present invention has the following remarkable advantages: considering the influence of real-time electricity price on user demand response, improving the participation of users in the electricity market, and increasing the flexibility and economy of the electricity market; under different demand response models, the energy storage response ability based on the stepped price demand response model is the strongest, which has an enhancing effect on the operation income of the active distribution network and the peak shaving and valley filling rate.

Description

Technical Field

[0001] The present invention relates to a method for optimizing the operation of a double-layer energy storage, and particularly to a method for optimizing the operation of a double-layer energy storage in a distribution network based on stepped price demand response, belonging to the field of power systems. Background Art

[0002] With the rapid development of demand-side management technology, the active load resources on the user side have continuously entered the power market. In order to utilize and manage these load resources, power operators on the user side generally carry out demand-side management through the modes of price guidance and contract management.

[0003] Traditional demand response research is based on the demand price elasticity coefficient in economics and is difficult to accurately reflect the response behavior of users. In the prior art, a driving demand response model and method based on cost-benefit analysis, a stochastic optimization model of a stepped demand response incentive mechanism considering the strong uncertainty of user response behavior, and a user response model that divides peak and valley periods by using an improved membership function and establishes an integrated time-of-use electricity price and incentive mechanism have been proposed. These demand response models lack the analysis of the impact of real-time electricity price on user demand response, resulting in a decrease in the degree of user participation in the power market and a reduction in the flexibility and economy of the power market. Summary of the Invention

[0004] In order to solve the deficiencies of the above technologies, the present invention provides a method for optimizing the operation of a double-layer energy storage in a distribution network based on stepped price demand response, so as to achieve the goal that the upper layer starts from the demand side, with the minimum user dissatisfaction, the maximum peak shaving and valley filling rate, the sum of the saved electricity costs and demand response subsidies being the largest, and the lower layer manages the combined power generation system of wind power, photovoltaics, and energy storage, and optimizes with the goal of maximizing the operation income of the power generation side distribution network.

[0005] In order to solve the above technical problems, the technical solution adopted by the present invention is: a method for optimizing the operation of a double-layer energy storage in a distribution network based on stepped price demand response, including the following steps:

[0006] S1. Considering the uncertainty of real-time electricity price fluctuations, an autoregressive moving average time series model and an interval optimization method are used to establish a dynamic real-time electricity price model;

[0007] S2. By analyzing the relationship between electricity price and demand response, a stepped price-demand response model is established;

[0008] S3. Combining the charge and discharge capacity and efficiency constraints of battery energy storage, a double-layer optimization model of a stepped price type demand response mechanism and regional energy storage operation is established, and optimization is carried out with the goal of maximizing the operation income of the power generation side distribution network.

[0009] Preferably, in step S1, the uncertainty is represented by the predicted value of the real-time electricity price and the price range of the prediction range result, and the established real-time electricity price model is as follows:

[0010]

[0011] p t,low ≤p t,unk ≤p t,high (2)

[0012] In the formula, p t represents the real-time electricity price at time t, r0 is a constant, h n is the autoregressive coefficient fitted through historical electricity price data, p t,unk is the electricity price fluctuation error value, p t,low is the lower limit model of the interval optimization model, p t,high is the upper limit model of the interval optimization model, P t-n is the real-time electricity price at time t - n.

[0013] Preferably, in step S2, different price intervals are determined. Taking the real-time electricity price as a known quantity and the load quantity participating in demand response as an unknown measured quantity, there is a functional relationship between them. The function of constructing the load quantity Δq j participating in demand response at time j is:

[0014]

[0015] In the formula, Δq j represents the load quantity participating in demand response at time j, Δp n is the length of the electricity price interval, k n is the load demand response coefficient of the nth section interval, δ i-1 and δ i are the left boundary value and right boundary value of the ith section interval respectively, p j is the electricity price at time j, p i is the electricity price of the ith interval node, δ0 and δ n are the upper limit and lower limit of the growth ratio of the real-time electricity price compared with the lowest electricity price predicted the day before, a and b are the upper and lower limits of the load quantity participating in demand response, n is the number of intervals, p min is the lowest electricity price.

[0016] Preferably, in step S2, considering the coordination of regional energy storage operation and demand response strategy, analyzing the source-network-load-storage supply-demand balance relationship in the active distribution network, the following multi-source and energy storage collaborative interaction constraints considering load demand response are constructed:

[0017] P j,ESS +P j,WT +P j,PV+P j,grid -P loss =P j,load -Δq j (4)

[0018] where P j,ESS is the electric power of the energy storage system at time j, P j,WT and P j,PV are the electric powers generated by wind power generation and photovoltaic power generation at time j respectively, P j,grid is the main grid power supply at time j, P j,load is the load power at time j, and P loss is the power loss in power transmission.

[0019] Preferably, in step S3, starting from the upper-layer demand side, with the minimum user dissatisfaction, the maximum peak shaving and valley filling rate, the sum of the saved electricity costs and the demand response subsidies being the largest, the regional energy storage is optimized based on the established stepped price-demand response model, and the lower layer manages the combined power generation system of wind power, photovoltaic power, and energy storage, and is optimized with the goal of maximizing the operating income of the power distribution network on the power generation side.

[0020] Preferably, in step S3, the objective function of the constructed two-layer optimization model of the energy storage is as follows:

[0021]

[0022] In the formula, ω1, ω2, ω3, and ω4 are the comprehensive evaluation weight values of user dissatisfaction and peak shaving and valley filling rate for different dimensions, ρ j，dis is the user dissatisfaction at time j, τ is the peak shaving and valley filling rate, C ele is the electricity cost saved by participating in the demand response, and C sub is the load subsidy income for participating in the demand response; the operating income of the power distribution network includes the operating income C ES of the energy storage system with low storage and high generation, the operating income C ren of renewable energy consumption, and the low-carbon operation loss C LC .

[0023] Preferably, in the upper-layer objective function of step S3:

[0024] The user dissatisfaction ρ j,dis at time j is expressed by the following formula:

[0025]

[0026] The peak shaving and valley filling rate τ is expressed by the following formula:

[0027]

[0028] The electricity cost C saved by participating in the demand responseele and the revenue C of the demand response load subsidy participation sub are expressed by the following formulas respectively:

[0029] C ele = p t ·Δq j (8)

[0030] C sub = μ sub ·Δq j (9)

[0031] where μ sub is the demand response subsidy coefficient.

[0032] Preferably, in the lower-level objective function of step S3:

[0033] The revenue C of the energy storage system operating with low storage and high discharge ES is expressed by the following formula:

[0034]

[0035] where p j is the real-time electricity price, P d is the discharge power of the energy storage system, P c is the charging power of the energy storage system, η d is the discharge efficiency, η c is the charging efficiency, t1 and t2 are the charging times, and t3 and t4 are the discharge times;

[0036] Using the electricity generated by distributed power sources in the distribution network to evaluate the renewable energy consumption benefit C ren , the calculation formula is as follows:

[0037]

[0038] where P j,ren is the electric power generated by renewable energy at time j, P j,WT and P j,PV are the electric powers generated by wind power and photovoltaic power at time j, N WT and N PV are the numbers of connected wind power and photovoltaic power respectively;

[0039] The low-carbon operation loss C of the power system LC is expressed as:

[0040] C LC = λ·ξ grid ·P j,grid (13)

[0041] where λ is the loss coefficient participating in carbon emissions, ξgrid is the carbon emission factor for main grid power generation.

[0042] Preferably, the upper-layer objective function constraint is:

[0043] Δq j,min ≤Δq j ≤Δq j,max (14)

[0044] Δp min ≤Δp n ≤Δp max (15)

[0045] k min ≤k n ≤k max (16)

[0046] where, Δq j,min and Δq j,max are respectively the minimum and maximum load amounts participating in demand response, Δp min and Δp max are respectively the minimum and maximum lengths of the decision electricity price interval, k min and k max are respectively the minimum and maximum load demand response coefficients.

[0047] Preferably, the lower-layer objective function constraint:

[0048] E ESS,min ≤E ESS,t ≤E ESS,max (17)

[0049] E ESS,t =E ESS,t-1 +Δt·(P d η d -P c η c ) (18)

[0050] P c,min ≤P c ≤P c,max (19)

[0051] P d,min ≤P d ≤P d,max (20)

[0052] where, E ESS,t is the capacity of the energy storage device at time t, E ESS,max and E ESS,min are respectively the maximum and minimum capacities of the energy storage device, P c,max and P c,minare the maximum and minimum values ​​of charging power, P d,max and P d,min are the maximum and minimum values ​​of the discharge power respectively;

[0053] P WT,min ≤P j,WT ≤P WT,max (twenty one)

[0054] P PV,min ≤P j,PV ≤P PV,max (twenty two)

[0055] P grid,min ≤P j,grid ≤P grid,max (twenty three)

[0056] Among them, P WT,max and P WT,min is the maximum power and minimum power produced by wind power generation at time t; P PV,max and P PV,min is the maximum power and minimum power produced by photovoltaic power generation at time t, P grid,max and P grid,min They are the maximum power and minimum power of the main grid power supply respectively.

[0057] Compared with the traditional price-based demand response mechanism, the present invention has the following significant advantages:

[0058] 1) Considering the impact of real-time electricity prices on user demand response, improving user participation in the electricity market, and increasing the flexibility and economy of the electricity market;

[0059] 2) Under different demand response models, the energy storage response capability based on the step-price demand response model is the strongest, which can improve the operating income of the active distribution network and the peak-shaving and valley-filling rate. BRIEF DESCRIPTION OF THE DRAWINGS

[0060] Figure 1 It is the overall flow chart of the present invention.

[0061] Figure 2 It is a real-time electricity price uncertainty curve diagram in the embodiment.

[0062] Figure 3 It is a step-price demand response load curve diagram with electricity price uncertainty in the embodiment.

[0063] Figure 4 It is a curve diagram of the changes of initial load, response load, photovoltaic power output and wind power output in the embodiment.

[0064] Figure 5It is a graph showing the optimization results of energy storage operation under the stepped price-based demand response mechanism model in the embodiment.

[0065] Figure 6 It is a comparison graph of the total system operation revenue and the peak shaving and valley filling rate under different demand response models in the embodiment. Specific implementation manners

[0066] The present invention will be further described in detail below in conjunction with the accompanying drawings and specific implementation manners.

[0067] As Figure 1 shown, the two-layer optimal operation method of distribution network energy storage based on stepped price demand response includes the following steps:

[0068] 1. Considering the uncertainty of real-time electricity price fluctuations, an autoregressive moving average time series model and interval optimization are used to establish a dynamic real-time electricity price model;

[0069] The fluctuations of real-time electricity prices directly affect the trading decisions of participants in the electricity market. In order to better formulate market strategies, it is necessary to predict future electricity prices. Based on the above considerations, the time series of historical electricity price data is analyzed, and an autoregressive moving average time series model and interval optimization are used to model the electricity price fluctuations. The uncertainty is represented by the predicted value and the price interval of the prediction range result of the real-time electricity price. The established real-time electricity price model is as follows:

[0070]

[0071] p t,low ≤ p t,unk ≤ p t,high (2)

[0072] In the formula, p t represents the real-time electricity price at time t, r0 is a constant, h n is the autoregressive coefficient fitted through historical electricity price data, p t,unk is the electricity price fluctuation error value, p t,low is the lower limit model of the interval optimization model, p t,high is the upper limit model of the interval optimization model, P t-n is the real-time electricity price at time t - n.

[0073] 2. By analyzing the relationship between electricity price and demand response, a stepped price-demand response model is established;

[0074] The load quantity participating in demand response has different changing trends at different time periods and electricity prices. Different price intervals are formulated. Taking the real-time electricity price as a known quantity and the load quantity participating in demand response as an unknown measured quantity, there is a certain functional relationship between them. The load quantity Δq participating in load demand response at time j is constructedj The function is as follows:

[0075]

[0076] In the formula, Δq j represents the load amount participating in demand response at time j. Δp n is the length of the electricity price interval, k n is the load demand response coefficient of the nth interval, δ i-1 and δ i are the left boundary value and right boundary value of the ith interval respectively. p j is the electricity price at time j, p i is the electricity price of the ith interval node, δ0 and δ n are the upper and lower limits of the growth ratio of the real-time electricity price compared with the lowest electricity price predicted the day before, a and b are the upper and lower limits of the load amount participating in demand response, n is the number of intervals, p min is the lowest electricity price.

[0077] Considering the coordinated operation of regional energy storage and demand response strategies, analyzing the source-network-load-storage supply-demand balance relationship in the active distribution network, the multi-source and energy storage collaborative interaction constraints (i.e., power balance constraint conditions) of the active distribution network considering load demand response are constructed as follows:

[0078] P j,ESS +P j,WT +P j,PV +P j,grid -P loss =P j,load -Δq j (4)

[0079] Among them, P j,ESS is the electric power of the energy storage system at time j, P j,WT and P j,PV are the electric powers generated by wind power generation and photovoltaic power generation at time j respectively, P j,grid is the main grid power supply at time j, P j,load is the load power at time j, P loss is the power loss in the power transmission.

[0080] 3. Combining the constraints such as the charge-discharge capacity and efficiency of the battery energy storage, a two-layer optimization model of the stepped price-based demand response mechanism and the operation of the regional energy storage is established.

[0081] Starting from the upper-level demand side, with the minimum user dissatisfaction, the maximum peak shaving and valley filling rate, the sum of the saved electricity costs and the demand response subsidies being the largest, the regional energy storage is optimized on the basis of the formulated stepped price-demand response model. The lower layer manages the combined power generation system of wind power, photovoltaics and energy storage, and is optimized with the goal of maximizing the operating income of the power distribution network on the power generation side;

[0082] The objective function of the constructed energy storage double - layer optimization model is as follows:

[0083]

[0084] In the formula, ω1, ω2, ω3, and ω4 are the comprehensive evaluation weight values of user dissatisfaction and peak - shaving and valley - filling rates for different dimensions, C ele is the electricity cost saved by participating in demand response, C sub is the load subsidy income from participating in demand response; the distribution network operation income includes the operation income C ES from the energy storage system charging at low levels and discharging at high levels, the operation income C ren from the consumption of renewable energy, and the low - carbon operation loss C LC . ρ j，dis is the user dissatisfaction at time j, and τ is the peak - shaving and valley - filling rate.

[0085] The user dissatisfaction changes quadratically with the change in energy demand. Therefore, the user dissatisfaction ρ j,dis at time j can be expressed by the following formula:

[0086]

[0087] Since users participate in load demand response, changing the inherent electricity consumption pattern, causing load reduction and transfer in time and space, thereby reducing the load peak - valley difference and smoothing the load curve. Therefore, the peak - shaving and valley - filling rate τ can be expressed by the following formula:

[0088]

[0089] The electricity cost C ele saved by participating in demand response and the load subsidy income C sub from participating in demand response are respectively as follows:

[0090] C ele = p t ·Δq j (8)

[0091] C sub = μ sub ·Δq j (9)

[0092] In formula (9), μ sub is the demand response subsidy coefficient.

[0093] The energy storage system charges during the low - electricity - price off - peak period and sells electric energy during the high - electricity - price peak period, achieving the purpose of increasing income through the electricity price difference. Therefore, the operation income C ES from the energy storage system charging at low levels and discharging at high levels is:

[0094]

[0095] Among them, p j is the real-time electricity price, P d is the discharge power of the energy storage system, P c is the charging power of the energy storage system, η d is the discharge efficiency, η c is the charging efficiency. t1 and t2 are the charging times, and t3 and t4 are the discharging times.

[0096] The electricity generated by distributed power sources in the distribution network is used to evaluate the renewable energy consumption benefit, and the calculation is as follows:

[0097]

[0098] Among them, P j,ren is the electric power generated by renewable energy at time j, P j,WT and P j,PV are the electric powers generated by wind power generation and photovoltaic power generation at time j, N WT , N PV are the numbers of connected wind power generation and photovoltaic power generation respectively.

[0099] The low-carbon operation loss of the power system can be expressed as:

[0100] C LC =λ·ξ grid ·P j,grid (13)

[0101] In formula (13), λ is the loss coefficient participating in carbon emissions, and ξ grid is the carbon emission coefficient of main grid power generation.

[0102] The upper-layer objective function constraint is:

[0103] Δq j,min ≤Δq j ≤Δq j,max (14)

[0104] Δp min ≤Δp n ≤Δp max (15)

[0105] k min ≤k n ≤k max (16)

[0106] Among them, Δq j,min and Δq j,max are the minimum and maximum load amounts participating in demand response respectively.

[0107] Δp min and Δp max are the minimum and maximum values of the length of the decision electricity price interval respectively. k min and k max are the minimum and maximum values of the load demand response coefficient respectively.

[0108] Lower layer objective function constraints:

[0109] E ESS,min ≤E ESS,t ≤E ESS,max (17)

[0110] E ESS,t =E ESS,t-1 +Δt·(P d η d -P c η c ) (18)

[0111] P c,min ≤P c ≤P c,max (19)

[0112] P d,min ≤P d ≤P d,max (20)

[0113] Among them, E ESS,t is the capacity of the energy storage device at time t, E ESS,max and E ESS,min are the maximum and minimum capacities of the energy storage device respectively, P c,max and P c,min are the maximum and minimum values of the charging power respectively, P d,max and P d,min are the maximum and minimum values of the discharging power respectively.

[0114] P WT,min ≤P j,WT ≤P WT,max (21)

[0115] P PV,min ≤P j,PV ≤P PV,max (22)

[0116] P grid,min ≤P j,grid ≤P grid,max (23)

[0117] Among them, P WT,max and P WT,min are the maximum and minimum powers generated by wind power at time t; P PV,max and PPV,min The maximum power and minimum power generated by photovoltaic power generation at time t. P grid,max and P grid,min are the maximum power and minimum power supplied by the main grid respectively.

[0118] The specific embodiments are as follows:

[0119] Based on the IEEE 14-node distribution network, a wind power generation system and a battery energy storage system are installed and configured at node 7, and a photovoltaic power generation system and a battery energy storage system are installed and configured at node 9. The capacity of the energy storage system is 2000 kWh; the capacities of the wind farm and the photovoltaic power station are both 7.5 MW; the fluctuation coefficients of wind power and photovoltaic power are 0.05 and 0.10 respectively.

[0120] Set the parameters for example verification. For the battery energy storage system, η d = η c = 0.95. For the wind power generation unit, N WT = 10. For the photovoltaic power generation unit, N PV = 10. For the load shedding loss ρ d = ρ r = 0.147 $. β t = 0.05. For the operation constraints, SOC max = 0.9, SOC min = 0.2.

[0121] P grid,min = 0 kW, P grid,max = 280 kW.

[0122] Considering the predicted value of the real-time electricity price and the price interval of the predicted range result to represent its uncertainty, analyze the time series of historical electricity price data, and use the autoregressive moving average time series model and interval optimization to model the electricity price fluctuation, and obtain the real-time electricity price uncertainty curve, as Figure 2 shown. Noon and evening are peak electricity consumption periods, so the predicted electricity price is relatively high.

[0123] Solve the optimal load demand response incentive coefficient of the price-based demand response mechanism through the particle swarm optimization algorithm, that is, k in formula (3) n . When considering the interval division of the real-time electricity price, the optimal step interval division of the stepped price-based demand response mechanism and the results of the load demand response incentive coefficient can be obtained, as shown in Table 1.

[0124] Table 1 Load demand response incentive coefficient

[0125]

[0126] When considering the interval division of real-time electricity prices, according to the obtained results of the optimal ladder interval division of the stepped price-based demand response mechanism and the load demand response incentive coefficient, the demand response load curve considering electricity price uncertainty is as follows Figure 3 As shown, after the initial load undergoes load demand response, the load is reduced and transferred. However, due to users' electricity consumption habits, only a small range of the load can be changed to a certain extent.

[0127] Figure 4 The electricity price curve in [reference] is the real-time electricity price data curve under the load demand response model based on the price elasticity coefficient matrix and the price-based demand response model without electricity price intervals. In addition, the output power curves of the wind power unit and the photovoltaic unit are provided. Noon and evening are peak electricity consumption periods, and the output power of the photovoltaic unit is relatively high, and the electricity price also increases accordingly.

[0128] As Figure 5 shown, the optimized operation results of the energy storage under the stepped price-based demand response mechanism model are obtained. It can be observed that in the time period of [0:00, 7:00], both the real-time electricity price and the electricity load are relatively low. Therefore, the energy storage system is in the charging state and the charging power is relatively large; in the time period of [3:00, 4:00], there is a small fluctuation in the electricity price, so the charging power of the energy storage decreases; in the time period of [7:00, 16:00], both the load level and the electricity price level are relatively high. To improve the total operation revenue of the distribution network, the power generation of photovoltaic and wind power is relatively high, and at the same time, the energy storage system discharges to increase the benefit of low storage and high discharge of the energy storage system; in the time period of [16:00, 18:00], due to the stable and decreasing real-time electricity price, the energy storage system charges again; in the time period of [18:00, 19:00], the electricity price increases slightly, so the energy storage system discharges at a relatively small power; in the time period of [19:00, 22:00], although it is a small peak of evening load, due to the large fluctuation of the electricity price, the energy storage system discharges at a relatively large power; finally, in the time period of [22:00, 24:00], as the real-time electricity price gradually decreases, the energy storage system charges and the charging power is relatively high. At the same time, due to the constraint that the SOC of the energy storage system remains unchanged at the beginning and end of the day, the energy storage system needs to charge.

[0129] The comparison results of the total system operation revenue and the peak shaving and valley filling rate under different demand response models are as Figure 6As shown, Scenario 1 is fixed load demand response, Scenario 2 is non-zonal price-based demand response, Scenario 3 is stepped price-based demand response with electricity price uncertainty, and Scenario 4 is stepped price-based demand response without electricity price uncertainty. The total optimal operation benefits of the active distribution network battery energy storage system for Scenarios 1 to 4 are $222.08, $246.7, $244.1, and $227.37 respectively. Due to the optimization of the flexible operation of the energy storage system and the relatively low low-carbon operation losses in Scenarios 2 and 3, the total benefits are slightly higher. In Scenarios 1 and 4, the increase in low-carbon operation losses results in the lowest total benefits, even though the energy storage system has slightly higher benefits when discharging less and generating more. The peak shaving and valley filling rates for Scenarios 1 to 4 are 4.95%, 8.37%, 11.52%, and 9.58% respectively. In Scenario 1, the load reduction and transfer amount affected by the load demand response based on the elasticity coefficient during high electricity load periods is relatively single, resulting in a lower discharge power of the energy storage system. Therefore, the peak shaving and valley filling rate of Scenario 1 is lower than that of the other three scenarios. In contrast, the peak shaving and valley filling rate of Scenario 3 based on the stepped price-based demand response model is the highest, verifying that the constructed model not only improves the total operation benefit of the active distribution network but also has the effect of smoothing the load curve and reducing the operation pressure of the power grid.

[0130] The above embodiments are not limitations to the present invention, and the present invention is not limited to the above examples either. Changes, modifications, additions, or substitutions made by those skilled in the art within the scope of the technical solution of the present invention also fall within the protection scope of the present invention.

Claims

1. A two-tier optimization operation method for energy storage in a distribution network based on step-by-step price demand response, characterized in that: The steps include: S1. Considering the uncertainty of real-time electricity price fluctuations, a dynamic real-time electricity price model is established using an autoregressive moving average time series model and interval optimization method; S2. Establish a step-by-step price-demand response model by analyzing the relationship between electricity price and demand response; Among them, different price ranges are proposed, with the real-time electricity price as the known quantity and the load participating in the demand response as the unknown quantity. There is a functional relationship between the two, and the load participating in the demand response at time j is constructed. j The function is: In the formula, Δq j represents the load participating in demand response at time j, Δp n is the length of the electricity price interval, k n is the load demand response coefficient of the nth interval, δ i-1 and δ i are the left and right boundary values ​​of the i-th interval, respectively, and p j is the electricity price at time j, p i is the electricity price of the node in the ith interval, δ0 and δ n is the upper and lower limits of the growth rate of the real-time electricity price compared to the lowest electricity price predicted a day ago, a and b are the upper and lower limits of the load participating in demand response, n is the number of intervals, and p min is the lowest electricity price; S3. Combining the battery energy storage charging and discharging capacity and efficiency constraints, a two-layer optimization model of the stepped price demand response mechanism and regional energy storage operation is established, with the optimization goal of maximizing the operating profit of the distribution network on the power generation side.

2. The method for double-layer optimization operation of energy storage in distribution network based on step-by-step price demand response according to claim 1 is characterized in that: In step S1, the uncertainty of the predicted value of the real-time electricity price and the price range of the predicted range result are used to represent the uncertainty, and the real-time electricity price model is established as follows: p t,low ≤p t,unk ≤p t,high (2) In the formula, p t represents the real-time electricity price at time t, r0 is a constant, h n is the autoregressive coefficient fitted by historical electricity price data, p t,unk is the price fluctuation error value, p t,low is the lower limit model of the interval optimization model, p t,high is the upper limit model of the interval optimization model, P t-n is the real-time electricity price at time tn.

3. The method for double-layer optimization operation of energy storage in distribution network based on step-by-step price demand response according to claim 1 is characterized in that: In step S2, considering the coordination between regional energy storage operation and demand response strategy, the source-grid-load-storage supply and demand balance relationship in the active distribution network is analyzed, and the multi-source and energy storage collaborative interaction constraints of the active distribution network considering load demand response are constructed as follows: P j,ESS +P j,WT +P j,PV +P j,grid -P loss =P j,load -Δq j (4) Among them, P j,ESS is the electric power of the energy storage system at time j, P j,WT and P j,PV are the electric power generated by wind power generation and photovoltaic power generation at time j, P j,grid is the main grid power supply at time j, P j,load is the load power at time j, P loss It is the power loss in electric energy transmission.

4. The method for double-layer optimization operation of energy storage in distribution network based on step-by-step price demand response according to claim 1 is characterized in that: In step S3, starting from the upper demand side, the regional energy storage is optimized on the basis of the established stepped price-demand response model in order to minimize user dissatisfaction, maximize the peak shaving and valley filling rate, maximize the sum of electricity cost savings and demand response subsidies, and manage the wind, photovoltaic, and energy storage combined power generation systems at the lower level, with the goal of maximizing the operating benefits of the distribution network on the power generation side.

5. The method for double-layer optimization operation of energy storage in distribution network based on step-by-step price demand response according to claim 4 is characterized in that: In step S3, the objective function of the constructed energy storage double-layer optimization model is as follows: Where ω1, ω2, ω3 and ω4 are the comprehensive evaluation weights for different dimensions of user dissatisfaction and peak-shaving and valley-filling rates, ρ j，dis is the user dissatisfaction at time j, τ is the peak shaving and valley filling rate, C ele The electricity cost saved by participating in demand response, C sub The income from demand response load subsidy; the income from distribution network operation includes the income from low storage and high generation operation of energy storage system C ES , Renewable energy consumption and operation income C ren , low carbon operation loss C LC .

6. The method for double-layer optimization operation of energy storage in distribution network based on step-by-step price demand response according to claim 5 is characterized in that: In the upper objective function of step S3: User dissatisfaction ρ at time j j,dis Expressed as follows: The peak shaving and valley filling rate τ is expressed by the following formula: Electricity cost savings from participating in demand response ele and the income from demand response load subsidy C sub They are expressed by the following formulas: C ele =p t ·Δq j (8) C sub =μ sub ·Δq j (9) Among them, μ sub is the demand response subsidy coefficient.

7. The method for double-layer optimization operation of energy storage in distribution network based on step-by-step price demand response according to claim 5 is characterized in that: In the lower objective function of step S3: Energy storage low storage high generation operation benefit C ES Expressed as follows: Among them, p j is the real-time electricity price, P d is the discharge power of the energy storage system, P c is the charging power of the energy storage system, η d is the discharge efficiency, η c is the charging efficiency, t1 and t2 are the charging time, and t3 and t4 are the discharging time; The electricity generated by distributed power generation in the distribution network is used to evaluate the benefits of renewable energy consumption C ren , the calculation formula is as follows: Among them, P j,ren is the electric power generated by renewable energy at time j, P j,WT and P j,PV is the electric power generated by wind power generation and photovoltaic power generation at time j, N WT 、N PV are the amounts of connected wind power and photovoltaic power, respectively; Low-carbon operation loss of power system C LC It is expressed as: C LC =l·x grid ·P j,grid (13) Among them, λ is the loss coefficient involved in carbon emission, ξ grid Carbon emission coefficient for main grid power generation.

8. The method for double-layer optimization operation of energy storage in distribution network based on step-by-step price demand response according to claim 6 is characterized in that: The upper objective function constraint is: Δq j,min ≤Δq j ≤Δq j,max (14) Δp min ≤Δp n ≤Δp max (15) k min ≤k n ≤k max (16) Among them, Δq j,min and Δq j,max are the minimum and maximum loads participating in demand response, Δp min and Δp max are the minimum and maximum lengths of the decision-making electricity price interval, respectively, and k min and k max are the minimum and maximum values ​​of the load demand response coefficient respectively.

9. The method for double-layer optimization operation of energy storage in distribution network based on step-by-step price demand response according to claim 7 is characterized in that: The lower objective function constraints are: AND ESS,min ≤E ESS,t ≤E ESS,max (17) E ESS,t =E ESS,t-1 +Δt·(P d or d -P c or c ) (18) P c,min ≤P c ≤P c,max (19) P d,min ≤P d ≤P d,max (20) Among them, E ESS,t is the capacity of the energy storage device at time t, E ESS,max and E ESS,min are the maximum and minimum capacities of the energy storage device, respectively, c,max and P c,min are the maximum and minimum values ​​of charging power, P d,max and P d,min are the maximum and minimum values ​​of the discharge power respectively; P WT,min ≤P j,WT ≤P WT,max (21) P PV,min ≤P j,PV ≤P PV,max (22) P grid,min ≤P j,grid ≤P grid,max (23) Among them, P WT,max and P WT,min is the maximum power and minimum power produced by wind power generation at time t; P PV,max and P PV,min is the maximum power and minimum power produced by photovoltaic power generation at time t, P grid,max and P grid,min They are the maximum power and minimum power of the main grid power supply respectively.

Images