Shared hybrid energy storage regulation and control strategy considering source-load-storage multi-time scale collaboration

Abstract

The invention relates to a shared hybrid energy storage regulation and control strategy considering source-load storage multi-time scale collaboration, and the strategy comprises the steps: obtaining a lease scale and an operation strategy of pumped storage resources according to source-load day-ahead prediction data by taking the maximum net benefit of pumped storage lease and the minimum hour-level fluctuation of a power grid power purchase curve as an optimization target; according to source load prediction data obtained through intra-day rolling, the maximum net benefit of storage battery renting and the minimum wind and light abandoning serve as optimization targets, and the renting scale and the operation strategy of storage battery resources in each stage are obtained; and obtaining the lease scale and the operation strategy of the hybrid energy storage resource in each stage by taking the maximum net benefit of the super capacitor and the minimum minute-level fluctuation ratio of the net load as optimization targets according to the source load ultra-short-term prediction data obtained in real time. According to the method, sufficient source consumption of renewable energy can be promoted, and meanwhile energy conservation and carbon reduction of the industrial park are promoted in the form of shared hybrid energy storage.

Description

Technical Field

[0001] The present invention relates to a shared hybrid energy storage control strategy that takes into account the multi-time scale coordination of sources, loads and storage, and belongs to the technical field of power system shared energy storage optimization. Background Art

[0002] Renewable energy generation will become an important part of the future power market. However, in this process, wind power and photovoltaic power have relatively poor direct grid-connected profitability due to their anti-peaking and uncertainty, which has a very adverse impact on the safety and reliability of the power grid and restricts the development of wind power and photovoltaic power in the power system.

[0003] Energy storage systems are widely used because they can effectively mitigate the adverse effects of wind power and photovoltaics on the power grid. Reasonable control strategies and methods are the key to their full effectiveness.

[0004] The accuracy of wind power and photovoltaic power prediction is closely related to its time scale. Long-time-scale power prediction is difficult to meet the requirements of precise control of energy storage. Although short-time-scale power prediction can meet the requirements of precise control of energy storage, short-time-scale energy storage control can only guarantee the optimal energy storage control within the current period, and cannot guarantee that the energy storage is in a reasonable normal working state for a long time. Summary of the invention

[0005] In order to solve the problems existing in the prior art, the present invention provides a shared hybrid energy storage system that introduces different types of shared hybrid energy storage systems at multiple time scales, thereby achieving a shared hybrid energy storage control strategy that considers the coordination of source, load and storage at multiple time scales to promote the consumption of renewable energy and energy conservation and carbon reduction in industrial parks.

[0006] In order to achieve the above-mentioned purpose, the technical solution proposed by the present invention is: a shared hybrid energy storage control strategy considering the multi-time scale coordination of source, load and storage, including:

[0007] For parks equipped with wind power and photovoltaic power, in the day-ahead stage, the day-ahead forecast data of park load and wind and solar output are obtained with a time span of 24 hours a day and a sampling interval of 1 hour. Based on the day-ahead forecast data, the objective function of pumped storage participating in the day-ahead stage is constructed with the optimization objectives of maximizing the net benefit of pumped storage leasing and minimizing the hourly fluctuation of the power grid purchase curve, and the optimal leasing scale and operation strategy of pumped storage resources are obtained.

[0008] In the intraday stage, the whole day is divided into four optimization stages, and the intraday forecast data of the park load and wind and solar output are obtained in a rolling manner with a time span of 6 hours and a sampling time interval of 15 minutes. Based on the intraday forecast data, the objective function of the battery participation in the intraday rolling optimization is constructed with the optimization goals of maximizing the net benefit of battery leasing and minimizing the abandonment of wind and solar power, and the optimal battery resource leasing scale and operation strategy in each optimization stage are obtained.

[0009] In the real-time stage, the ultra-short-term forecast data of the park load and wind and solar output are obtained with a time span of 15 minutes and a sampling time interval of 1 minute. Based on the ultra-short-term forecast data, the objective function of supercapacitor participation in real-time correction is constructed with the optimization goals of maximizing the net benefit of supercapacitors and minimizing the minute-level fluctuation rate of net load, and the optimal leasing scale and operation strategy of hybrid energy storage resources in each real-time stage are obtained.

[0010] The leasing scale and operating strategy of the optimal pumped storage resources, the leasing scale and operating strategy of the optimal battery resources in each optimization stage, and the leasing scale and operating strategy of the optimal hybrid energy storage resources in each stage are integrated on a time scale to obtain a shared hybrid energy storage regulation strategy.

[0011] The further design of the above technical solution is as follows: the objective function f1 of the pumped storage participating in the day-ahead stage is:

[0012] maxf1=(R PS -H PS -C grid )-α1Flu1

[0013]

[0014] H PS =h PS,P (P PS,Pmax +P PS,Gmax )+h PS,V (V u,max +V d,max )

[0015]

[0016] Among them, R PS represents the electricity price arbitrage benefit of pumped storage in the day-ahead planning stage; H PS represents the rental cost of pumped storage in one day; C grid represents the electricity purchase cost that the industrial park needs to pay from the power grid after configuring large-capacity pumped storage; Flu1 represents the hourly volatility of the power purchase curve of the power grid, α1 represents the hourly volatility weighting coefficient; t1 represents the index variable of the day-ahead stage, t1=1,2,…,24, Δt1 represents the optimization decision time interval; ppv (t1) represents the peak and valley time-of-use electricity price at time t1; P PS,G (t1) and P PS,P (t1) represents the generating power and pumping power of the pumped storage at time t1 respectively; h PS,P and h PS,V They represent the unit power rental cost coefficient and unit capacity rental cost coefficient of pumped storage respectively; P PS,Gmax Indicates the maximum power generation of pumped storage; P PS,Pmax Indicates the maximum pumping power of pumped storage; V u,max Respectively represent the maximum capacity of the upper reservoir; V d,max Respectively represent the maximum capacity of the lower reservoir; P da,gird (t1) represents the purchased power of the industrial park at time t1 in the day-ahead optimization; E T Indicates the transformer capacity of the industrial park.

[0017] The constraint condition of the objective function f1 of the pumped storage participating in the day-ahead stage is:

[0018] (1) Power balance constraints in the day-ahead phase;

[0019] P da,grid (t1) = P da,load (t1)-P da,WP (t1)-P da,PV (t1)-P PS,G (t1)+P PS,P (t1)

[0020] Among them, P da,load (t1),P da,WP (t1),P da,PV (t1) represents the forecast data of industrial park power load, wind power output and photovoltaic output at time t1 in the day-ahead period;

[0021] (2) Physical constraints of pumped storage;

[0022] V u (t1) = V u (t1-1)-Q G (t1)+Q P (t1)

[0023] V d (t1) = V d (t1-1)-Q P (t1)+Q G (t1)

[0024] V u,min ≤V u (t1)≤V u,max

[0025] V d,min ≤V d (t1)≤V d,max

[0026] V u (1) = V u (twenty four)

[0027] V d (1) = V d (twenty four)

[0028] Q G (t1) = P PS,G (t1) / η PS,G ·ρ W ·g·h

[0029] Q P (t1) = η PS,P ·P PS,P (t1) / ρ W ·g·h

[0030] B PS,G (t1)·P PS,Gmin ≤P PS,G (t1)≤B PS,G (t1)·P PS,Gmax

[0031] B PS,P (t1)·P PS,Pmin ≤P PS,P (t1)≤B PS,P (t1)·P PS,Pmax

[0032] 0≤B PS,G (t1)+B PS,P (t1)≤1

[0033] Among them, V u (t1) and V d (t1) represents the storage capacity of the upper and lower reservoirs of the pumped storage at time t1; Q G (t1) and Q P (t1) respectively represent the power generation flow and pumping flow of the pumped storage at time t1; V u,min Represents the minimum capacity of the upper reservoir; V d,min represents the minimum capacity of the lower reservoir; η PS,G and η PS,S They represent the power generation efficiency and pumping efficiency of pumped storage respectively; ρ W represents water density; g represents gravitational acceleration; h represents the average water head height of pumped storage; B PS,G and B PS,Sis a Boolean variable, and state 1 indicates that the pumped storage is in the power generation state and the pumping state; P PS,Gmin represents the minimum generating power of pumped storage; P PS,Pmin Indicates the minimum pumping power of pumped storage.

[0034] The objective function f2 of the battery participating in the intraday rolling optimization is:

[0035] maxf2=(R BS -H BS )-α2C ab

[0036]

[0037] H BS =h BS,P ·P BS +h BS,E ·E BS

[0038]

[0039] Among them, R BS represents the electricity price arbitrage income of the battery in the rolling optimization stage; H BS Represents the rental cost of the battery; C ab represents the cost of wind and solar power abandonment; α2 represents the weighted coefficient of wind and solar power abandonment cost; t2 represents the time index variable of the rolling optimization stage, t2=1,2,…,24, Δt2 optimizes the decision time interval; p pv (t2) represents the peak-valley time-of-use electricity price at time t2; P BS ,dis (t2) and P BS,ch (t2) respectively represent the discharge power and charge power of the battery at time t2; h BS,P and h BS,E They represent the unit power rental cost coefficient and unit capacity rental cost coefficient of the battery respectively; P BS and E BS Respectively represent the rated power and rated capacity of the battery; ab represents the penalty coefficient for abandoning wind and solar power in the park; p WP represents the grid-connected price of wind power; p PV Represents the photovoltaic grid-connected price; P WP,ab (t2) represents the power abandoned by the wind power station at time t2; P PV,ab (t2) represents the power discarded by the photovoltaic power station at time t2.

[0040] The constraint condition of the objective function f2 of the battery participating in the intraday rolling optimization is:

[0041] (1) Power balance constraints during the intraday rolling phase;

[0042] P roll,load (t2)-P roll,grid (t2)-(P PS,G (t2)-P PS,P (t2))=

[0043] (P roll,WP (t2)-P WP,ab (t2))+(P roll,PV (t2)-P PV,ab (t2))+(P BS,dis (t2)-P BS,ch (t2))

[0044] Among them, P roll,load (t2), P roll,grid (t2), P roll,WP (t2), P roll,PV (t2) respectively represent the forecast data of industrial park power load, external power purchase, wind power output and photovoltaic output at time t2 in the intraday rolling optimization stage;

[0045] (2) Physical constraints on batteries;

[0046]

[0047] In the formula, SOC BS (t2) represents the state of charge of the battery at time t2; η BS,ch and η BS,dis Respectively represent the charging efficiency and discharging efficiency of the battery; SOC BS,min and SOC BS,max Respectively represent the minimum and maximum value of the battery charge state; P BS ,ch,max and η BS,dis,max Respectively represent the maximum charging power and maximum discharging power allowed by the battery; B BS,ch (t2) and B BS,dis (t2) are Boolean variables, and state 1 represents the charging state and discharging state of the battery in the t2 period respectively.

[0048] The objective function f3 of the supercapacitor participating in the real-time correction is:

[0049] maxf3=(R SC -H SC )-α3Flu3

[0050]

[0051] H SC=h SC,P ·P SC +h SC,E ·E SC

[0052]

[0053] Among them, R SC represents the electricity price arbitrage profit of the supercapacitor in the real-time correction stage; H SC represents the rental cost of the supercapacitor; Flu3 represents the minute-level volatility of the power grid purchase curve, α3 represents the minute-level volatility weighting coefficient; t3 represents the index variable of the day-ahead stage, t3=1,2,…,15, Δt3 optimizes the decision time interval; p pv (t3) represents the peak-valley time-of-use electricity price at time t3; P SC,dis (t3) and P SC,ch (t3) respectively represent the discharge power and charging power of the supercapacitor at time t3; h SC,P and h SC ,E They represent the unit power rental cost coefficient and unit capacity rental cost coefficient of the supercapacitor respectively; P SC and E SC Respectively represent the rated power and rated capacity of the supercapacitor; P real,grid (t3) represents the external power purchase amount at time t3, that is, the net load demand of the industrial park.

[0054] The constraint condition of the objective function f3 of the supercapacitor participating in the real-time correction is:

[0055] (1) Correct power balance constraints in real time;

[0056] P real,load (t3)-P real,grid (t3)-(P PS,G (t3)-P PS,P (t3))-(P BS,dis (t3)-P BS,ch (t3))=

[0057] (P real,WP (t3)-P WP,ab (t3))+(P real,PV (t3)-P PV,ab (t3))+(P SC,dis (t3)-P SC,ch (t3))

[0058] Among them, P real,load (t3), P real,grid (t3), P real,WP (t3), P real,PV(t3) represent the forecast data of industrial park power load, external power purchase, wind power output and photovoltaic power output at time t3 in the real-time correction optimization stage.

[0059] (2) Physical constraints of supercapacitors;

[0060]

[0061] In the formula, SOC SC (t3) represents the charge state of the supercapacitor at time t3; η SC,ch and η SC,dis Respectively represent the charging efficiency and discharging efficiency of supercapacitor; SOC SC,min and SOC SC,max Respectively represent the minimum and maximum value of the supercapacitor charge state; P SC,ch,max and η SC,dis,max Respectively represent the maximum charging power and maximum discharging power allowed by the supercapacitor; B SC,ch (t3) and B SC,dis (t3) are Boolean variables, and state 1 represents the charging state and discharging state of the supercapacitor in the t3 period respectively.

[0062] The beneficial effects of the present invention are:

[0063] (1) This paper takes industrial parks equipped with wind and solar power plants as the research object, and studies the leasing scale and operation strategy of shared hybrid energy storage from three time scales: day-ahead planning, intraday rolling adjustment, and real-time correction. The introduction of shared hybrid energy storage not only reduces the storage cost of industrial parks, but also increases the flexibility of electricity use in industrial parks.

[0064] (2) The present invention comprehensively considers the multi-time scale uncertainty of renewable energy and the demand savings of net load under the two-part electricity price background at multiple time scales, and improves the wind and solar power consumption ratio through three-stage refined adjustment. At the same time, the introduction of shared hybrid energy storage also reduces the overall external power purchase demand and net load curve volatility of the industrial park, which not only promotes the full source consumption of renewable energy, but also promotes energy conservation and carbon reduction in the industrial park through shared hybrid energy storage. BRIEF DESCRIPTION OF THE DRAWINGS

[0065] Figure 1 Flowchart of the invention strategy. DETAILED DESCRIPTION

[0066] The present invention is described in detail below with reference to the accompanying drawings and specific embodiments.

[0067] Example

[0068] This embodiment considers the shared hybrid energy storage leasing strategy of multi-time scale coordination of source, load and storage (combining the data of the power generation side and the load side to configure the energy storage), and is aimed at industrial parks equipped with a certain scale of wind power and photovoltaic power, and optimizes the leasing scale of shared hybrid energy storage at three time scales: day-ahead, intraday, and real-time.

[0069] Since pumped storage is a super-large capacity energy storage device, it is particularly suitable for meeting the base load demand of industrial parks; batteries have a faster regulation speed than pumped storage and are suitable for hourly intraday rolling optimization; supercapacitors, as a power-type energy storage, can be used to smooth minute-level fluctuations in the net load curve. Therefore, the shared hybrid energy storage resources in this embodiment correspond to three energy storage resources with different time scales, namely pumped storage, batteries and supercapacitors.

[0070] The strategy of this embodiment is as follows: Figure 1 As shown, the following steps are included:

[0071] In the day-ahead stage, with a time span of 24 hours and a sampling time interval of 1 hour, the day-ahead forecast data of the park load and wind and solar power output are obtained; based on the day-ahead forecast data, the objective function of pumped storage participating in the day-ahead stage is constructed with the maximum net benefit of pumped storage leasing and the minimum hourly fluctuation of the power grid purchase curve as the optimization goals, and the optimal leasing scale and operation strategy of pumped storage resources are obtained.

[0072] The objective function and constraints of pumped storage in the day-ahead phase are as follows:

[0073] (I) Objective function f1: maximizing the net benefit of pumped storage leasing in the day-ahead phase and minimizing the hourly fluctuation of the power grid purchase curve:

[0074] maxf1=(R PS -H PS -C grid )-α1Flu1

[0075]

[0076] H PS =h PS,P (P PS,Pmax +P PS,Gmax )+h PS,V (V u,max +V d,max )

[0077]

[0078] Among them, R PS represents the electricity price arbitrage benefit of pumped storage in the day-ahead planning stage; H PS represents the rental cost of pumped storage in one day; Cgrid represents the electricity purchase cost that the industrial park needs to pay from the power grid after configuring large-capacity pumped storage; Flu1 represents the hourly volatility of the power purchase curve of the power grid, α1 represents the weighted coefficient of the hourly volatility; t1 represents the index variable of the day-ahead stage, t1=1,2,…,24, Δt1 optimizes the decision time interval; p pv (t1) represents the peak and valley time-of-use electricity price at time t1; P PS ,G (t1) and P PS,P (t1) represents the generating power and pumping power of the pumped storage at time t1 respectively; h PS,P and h PS,V They represent the unit power rental cost coefficient and unit capacity rental cost coefficient of pumped storage respectively; P PS,Gmax Indicates the maximum power generation of pumped storage; P PS,Pmax Indicates the maximum pumping power of pumped storage; V u,max Respectively represent the maximum capacity of the upper reservoir; V d,max Respectively represent the maximum capacity of the lower reservoir; P da,gird (t1) represents the purchased power of the industrial park at time t1 in the day-ahead optimization; E T Indicates the transformer capacity of the industrial park.

[0079] (II) Constraints

[0080] (1) Power balance constraints in the day-ahead phase;

[0081] P da,grid (t1) = P da,load (t1)-P da,WP (t1)-P da,PV (t1)-P PS,G (t1)+P PS,P (t1)

[0082] Among them, P da,load (t1),P da,WP (t1),P da,PV (t1) respectively represent the forecast data of industrial park power load, wind power output and photovoltaic output at time t1 in the day-ahead period.

[0083] (2) Physical constraints of pumped storage;

[0084] V u (t1) = V u (t1-1)-Q G (t1)+Q P (t1)

[0085] V d (t1) = V d (t1-1)-QP (t1)+Q G (t1)

[0086] V u,min ≤V u (t1)≤V u,max

[0087] V d,min ≤V d (t1)≤V d,max

[0088] V u (1) = V u (twenty four)

[0089] V d (1) = V d (twenty four)

[0090] Q G (t1) = P PS,G (t1) / η PS,G ·ρ W ·g·h

[0091] Q P (t1) = η PS,P ·P PS,P (t1) / ρ W ·g·h

[0092] B PS,G (t1)·P PS,Gmin ≤P PS,G (t1)≤B PS,G (t1)·P PS,Gmax

[0093] B PS,P (t1)·P PS,Pmin ≤P PS,P (t1)≤B PS,P (t1)·P PS,Pmax

[0094] 0≤B PS,G (t1)+B PS,P (t1)≤1

[0095] Among them, V u (t1) and V d (t1) represents the storage capacity of the upper and lower reservoirs of the pumped storage at time t1; Q G (t1) and Q P (t1) respectively represent the power generation flow and pumping flow of the pumped storage at time t1; V u,min Represents the minimum capacity of the upper reservoir; V d,min represents the minimum capacity of the lower reservoir; η PS,Gand η PS,S They represent the power generation efficiency and pumping efficiency of pumped storage respectively; ρ W represents water density; g represents gravitational acceleration; h represents the average water head height of pumped storage; B PS,G and B PS,S is a Boolean variable, and state 1 indicates that the pumped storage is in the power generation state and the pumping state; P PS,Gmin represents the minimum generating power of pumped storage; P PS,Pmin Indicates the minimum pumping power of pumped storage.

[0096] In the intraday stage, the whole day is divided into four optimization stages, and the intraday forecast data of the park load and wind and solar power output are obtained in a rolling manner with a time span of 6 hours and a sampling time interval of 15 minutes. According to the intraday forecast data, the objective function of the battery participation in the intraday rolling optimization is constructed with the maximization of the net benefit of battery leasing and the minimization of wind and solar abandonment as the optimization goals, and the optimal battery resource leasing scale and operation strategy in each optimization stage are obtained.

[0097] (I) Objective function f2: maximizing the net benefit of battery leasing and minimizing wind and solar power abandonment in the rolling phase within the day

[0098] maxf2=(R BS -H BS )-α2C ab

[0099]

[0100] H BS =h BS,P ·P BS +h BS,E ·E BS

[0101]

[0102] Among them, R BS represents the electricity price arbitrage income of the battery in the rolling optimization stage; H BS Represents the rental cost of the battery; C ab represents the cost of wind and solar power abandonment; α2 represents the weighted coefficient of wind and solar power abandonment cost; t2 represents the time index variable of the rolling optimization stage, t2=1,2,…,24, Δt2 optimizes the decision time interval; p pv (t2) represents the peak-valley time-of-use electricity price at time t2; P BS ,dis (t2) and P BS,ch (t2) respectively represent the discharge power and charge power of the battery at time t2; h BS,P and h BS,EThey represent the unit power rental cost coefficient and unit capacity rental cost coefficient of the battery respectively; P BS and E BS Respectively represent the rated power and rated capacity of the battery; ab represents the penalty coefficient for abandoning wind and solar power in the park; p WP represents the grid-connected price of wind power; p PV Represents the photovoltaic grid-connected price; P WP,ab (t2) represents the power abandoned by the wind power station at time t2; P PV,ab (t2) represents the power discarded by the photovoltaic power station at time t2.

[0103] (II) Constraints

[0104] (1) Power balance constraints during the intraday rolling phase

[0105] Proll,load(t2)-Proll,grid(t2)-(PPS,G(t2)-PPS,P(t2))=

[0106] (P roll,WP (t2)-P WP,ab (t2))+(P roll,PV (t2)-P PV,ab (t2))+(P BS,dis (t2)-P BS,ch (t2))

[0107] Among them, P roll,load (t2), P roll,grid (t2), P roll,WP (t2), P roll,PV (t2) represent the forecast data of industrial park power load, external power purchase, wind power output and photovoltaic power output at time t2 in the intraday rolling optimization phase.

[0108] (2) Physical constraints of batteries

[0109]

[0110] In the formula, SOC BS (t2) represents the state of charge of the battery at time t2; η BS,ch and η BS,dis Respectively represent the charging efficiency and discharging efficiency of the battery; SOC BS,min and SOC BS,max Respectively represent the minimum and maximum value of the battery charge state; P BS ,ch,max and η BS,dis,max Respectively represent the maximum charging power and maximum discharging power allowed by the battery; B BS,ch(t2) and B BS,dis (t2) are Boolean variables, and state 1 represents the charging state and discharging state of the battery in the t2 period respectively.

[0111] In the real-time stage, with a time span of 15 minutes and a sampling time interval of 1 minute, ultra-short-term forecast data on the park load and wind and solar output are obtained; based on the ultra-short-term forecast data, the objective function for supercapacitor participation in real-time correction is constructed with the maximum net benefit of supercapacitors and the minimum minute-level volatility of net load as the optimization goals, and the leasing scale and operation strategy of the optimal hybrid energy storage resources in each real-time stage are obtained; in the real-time stage, a day is divided into spans of 15 minutes, for a total of 96 time periods.

[0112] Real-time correction in the real-time stage, taking the time period of 8:00-8:15 as an example:

[0113] ① Initial optimization: At 8:00, the system obtains ultra-short-term forecast data of load and wind and solar power output in the next 15 minutes. Based on the forecast data, the system performs an optimization to obtain the preliminary operation plan of the supercapacitor from 8:00 to 8:15, including the charging and discharging strategy;

[0114] ② Minute-by-minute sampling and real-time correction: Real-time data is collected every minute. For example, actual load and wind and solar data are collected at 8:01. Real-time data is compared with predicted data, load deviation is calculated, and the supercapacitor at 8:01 is adjusted immediately to ensure that the operation of the supercapacitor is closer to actual needs.

[0115] At the subsequent time points of 8:02, 8:03, etc., the above process is repeated to achieve point-by-point correction of the supercapacitor.

[0116] (I) Objective function f3: Maximizing the net benefit of supercapacitors and minimizing the minute-level fluctuation rate of net load

[0117] maxf3=(R SC -H SC )-α3Flu3

[0118]

[0119] H SC =h SC,P ·P SC +h SC,E ·E SC

[0120]

[0121] Among them, R SC represents the electricity price arbitrage profit of the supercapacitor in the real-time correction stage; H SCrepresents the rental cost of the supercapacitor; Flu3 represents the minute-level volatility of the power grid purchase curve, α3 represents the minute-level volatility weighting coefficient; t3 represents the index variable of the day-ahead stage, t3=1,2,…,15, Δt3 optimizes the decision time interval; p pv (t3) represents the peak-valley time-of-use electricity price at time t3; P SC,dis (t3) and P SC,ch (t3) respectively represent the discharge power and charging power of the supercapacitor at time t3; h SC,P and h SC ,E They represent the unit power rental cost coefficient and unit capacity rental cost coefficient of the supercapacitor respectively; P SC and E SC Respectively represent the rated power and rated capacity of the supercapacitor; P real,grid (t3) represents the external power purchase amount at time t3, that is, the net load demand of the industrial park.

[0122] (II) Constraints

[0123] (1) Real-time correction of power balance constraints

[0124] P real,load (t3)-P real,grid (t3)-(P PS,G (t3)-P PS,P (t3))-(P BS,dis (t3)-P BS,ch (t3))=

[0125] (P real,WP (t3)-P WP,ab (t3))+(P real,PV (t3)-P PV,ab (t3))+(P SC,dis (t3)-P SC,ch (t3))

[0126] Among them, P real,load (t3), P real,grid (t3), P real,WP (t3), P real,PV (t3) represent the forecast data of industrial park power load, external power purchase, wind power output and photovoltaic power output at time t3 in the real-time correction optimization stage.

[0127] (2) Physical constraints of supercapacitors

[0128]

[0129] In the formula, SOC SC (t3) represents the charge state of the supercapacitor at time t3; ηSC,ch and η SC,dis Respectively represent the charging efficiency and discharging efficiency of supercapacitor; SOC SC,min and SOC SC,max Respectively represent the minimum and maximum value of the supercapacitor charge state; P SC,ch,max and η SC,dis,max Respectively represent the maximum charging power and maximum discharging power allowed by the supercapacitor; B SC,ch (t3) and B SC,dis (t3) are Boolean variables, and state 1 represents the charging state and discharging state of the supercapacitor in the t3 period respectively.

[0130] Finally, the leasing scale and operation strategy of the optimal pumped storage resources, the leasing scale and operation strategy of the optimal battery resources in each optimization stage, and the leasing scale and operation strategy of the optimal hybrid energy storage resources in each stage are integrated on a time scale to obtain a shared hybrid energy storage regulation strategy.

[0131] The technical solutions of the present invention are not limited to the above-mentioned embodiments, and any technical solutions obtained by equivalent replacement methods fall within the scope of protection required by the present invention.

Claims

1. A shared hybrid energy storage control strategy considering the multi-time scale coordination of source, load and storage, characterized by: For parks equipped with wind power and photovoltaic power, in the day-ahead stage, the day-ahead forecast data of park load and wind and solar output are obtained with a time span of 24 hours a day and a sampling interval of 1 hour. Based on the day-ahead forecast data, the objective function of pumped storage participating in the day-ahead stage is constructed with the optimization objectives of maximizing the net benefit of pumped storage leasing and minimizing the hourly fluctuation of the power grid purchase curve, and the optimal leasing scale and operation strategy of pumped storage resources are obtained. In the intraday stage, the whole day is divided into four optimization stages, and the intraday forecast data of the park load and wind and solar output are obtained in a rolling manner with a time span of 6 hours and a sampling time interval of 15 minutes. Based on the intraday forecast data, the objective function of the battery participation in the intraday rolling optimization is constructed with the optimization goals of maximizing the net benefit of battery leasing and minimizing the abandonment of wind and solar power, and the optimal battery resource leasing scale and operation strategy in each optimization stage are obtained. In the real-time stage, the ultra-short-term forecast data of the park load and wind and solar output are obtained with a time span of 15 minutes and a sampling time interval of 1 minute. Based on the ultra-short-term forecast data, the objective function of supercapacitor participation in real-time correction is constructed with the optimization goals of maximizing the net benefit of supercapacitors and minimizing the minute-level fluctuation rate of net load, and the optimal leasing scale and operation strategy of hybrid energy storage resources in each real-time stage are obtained. The leasing scale and operating strategy of the optimal pumped storage resources, the leasing scale and operating strategy of the optimal battery resources in each optimization stage, and the leasing scale and operating strategy of the optimal hybrid energy storage resources in each stage are integrated on a time scale to obtain a shared hybrid energy storage regulation strategy.

2. The shared hybrid energy storage control strategy considering multi-time scale coordination of source, load and storage according to claim 1 is characterized by: The objective function f1 of the pumped storage participating in the day-ahead stage is: maxf1=(R PS -H PS -C grid )-α1Flu1 H PS =h PS,P (P PS,Pmax +P PS,Gmax )+h PS,V (V u,max +V d,max ) Among them, R PS represents the electricity price arbitrage benefit of pumped storage in the day-ahead planning stage; H PS represents the rental cost of pumped storage in one day; C grid represents the electricity purchase cost that the industrial park needs to pay from the power grid after configuring large-capacity pumped storage; Flu1 represents the hourly volatility of the power purchase curve of the power grid, α1 represents the hourly volatility weighting coefficient; t1 represents the index variable of the day-ahead stage, t1=1,2,…,24, Δt1 represents the optimization decision time interval; p pv (t1) represents the peak and valley time-of-use electricity price at time t1; P PS ,G (t1) and P PS,P (t1) represents the generating power and pumping power of the pumped storage at time t1 respectively; h PS,P and h PS,V They represent the unit power rental cost coefficient and unit capacity rental cost coefficient of pumped storage respectively; P PS,Gmax Indicates the maximum power generation of pumped storage; P PS,Pmax Indicates the maximum pumping power of pumped storage; V u,max Respectively represent the maximum capacity of the upper reservoir; V d,max Respectively represent the maximum capacity of the lower reservoir; P da,gird (t1) represents the purchased power of the industrial park at time t1 in the day-ahead optimization; E T Indicates the transformer capacity of the industrial park.

3. The shared hybrid energy storage control strategy considering multi-time scale coordination of source, load and storage according to claim 2 is characterized by: The constraint condition of the objective function f1 of the pumped storage participating in the day-ahead stage is: (1) Power balance constraints in the day-ahead phase; P da,grid (t1)=P da,load (t1)-P da,WP (t1)-P da,PV (t1)-P PS,G (t1)+P PS,P (t1) Among them, P da,load (t1),P da,WP (t1),P da,PV (t1) represents the forecast data of industrial park power load, wind power output and photovoltaic output at time t1 in the day-ahead period; (2) Physical constraints of pumped storage; 5 u (t1)=V u (t1-1)-Q G (t1)+Q P (t1) 5 d (t1)=V d (t1-1)-Q P (t1)+Q G (t1) 5 u,min ≤V u (t1)≤V u,max 5 d,min ≤V d (t1)≤V d,max V u (1)=V u (24) V d (1)=V d (24) Q G (t1)=P PS,G (t1) / η PS,G ·ρ W ·g·h Q P (t1)=η PS,P ·P PS,P (t1) / ρ W ·g·h B PS,G (t1)·P PS,Gmin ≤P PS,G (t1)≤B PS,G (t1)·P PS,Gmax B PS,P (t1)·P PS,Pmin ≤P PS,P (t1)≤B PS,P (t1)·P PS,Pmax 0≤B PS,G (t1)+B PS,P (t1)≤1 Among them, V u (t1) and V d (t1) represents the storage capacity of the upper and lower reservoirs of the pumped storage at time t1; Q G (t1) and Q P (t1) respectively represent the power generation flow and pumping flow of the pumped storage at time t1; V u,min Represents the minimum capacity of the upper reservoir; V d,min represents the minimum capacity of the lower reservoir; η PS,G and η PS,S They represent the power generation efficiency and pumping efficiency of pumped storage respectively; ρ W represents water density; g represents gravitational acceleration; h represents the average water head height of pumped storage; B PS,G and B PS,S is a Boolean variable, and state 1 indicates that the pumped storage is in the power generation state and the pumping state; P PS,Gmin represents the minimum generating power of pumped storage; P PS,Pmin Indicates the minimum pumping power of pumped storage.

4. The shared hybrid energy storage control strategy considering multi-time scale coordination of source, load and storage according to claim 1 is characterized by: The objective function f2 of the battery participating in the intraday rolling optimization is: maxf2=(R BS -H BS )-α2C ab H BS =h BS,P ·P BS +h BS,E ·E BS Among them, R BS represents the electricity price arbitrage income of the battery in the rolling optimization stage; H BS Represents the rental cost of the battery; C ab represents the cost of wind and solar power abandonment; α2 represents the weighted coefficient of wind and solar power abandonment cost; t2 represents the time index variable of the rolling optimization stage, t2=1,2,…,24, Δt2 optimizes the decision time interval; p pv (t2) represents the peak-valley time-of-use electricity price at time t2; P BS,dis (t2) and P BS,ch (t2) respectively represent the discharge power and charge power of the battery at time t2; h BS,P and h BS,E They represent the unit power rental cost coefficient and unit capacity rental cost coefficient of the battery respectively; P BS and E BS Respectively represent the rated power and rated capacity of the battery; ab represents the penalty coefficient for abandoning wind and solar power in the park; p WP represents the grid-connected price of wind power; p PV Represents the photovoltaic grid-connected price; P WP,ab (t2) represents the power abandoned by the wind power station at time t2; P PV,ab (t2) represents the power discarded by the photovoltaic power station at time t2.

5. The shared hybrid energy storage control strategy considering multi-time scale coordination of source, load and storage according to claim 4 is characterized by: The constraint condition of the objective function f2 of the battery participating in the intraday rolling optimization is: (1) Power balance constraints during the intraday rolling phase; P roll,load (t2)-P roll,grid (t2)-(P PS,G (t2)-P PS,P (t2))=(P roll,WP (t2)-P WP,ab (t2))+(P roll,PV (t2)-P PV,ab (t2))+(P BS,dis (t2)-P BS,ch (t2)) Among them, P roll,load (t2), P roll,grid (t2), P roll,WP (t2), P roll,PV (t2) respectively represent the forecast data of industrial park power load, external power purchase, wind power output and photovoltaic output at time t2 in the intraday rolling optimization stage; (2) Physical constraints on batteries; In the formula, SOC BS (t2) represents the state of charge of the battery at time t2; η BS,ch and η BS,dis Respectively represent the charging efficiency and discharging efficiency of the battery; SOC BS,min and SOC BS,max Respectively represent the minimum and maximum value of the battery charge state; P BS,ch,max and η BS,dis,max Respectively represent the maximum charging power and maximum discharging power allowed by the battery; B BS,ch (t2) and B BS,dis (t2) are Boolean variables, and state 1 represents the charging state and discharging state of the battery in the t2 period respectively.

6. The shared hybrid energy storage control strategy considering multi-time scale coordination of source, load and storage according to claim 1 is characterized by: The objective function f3 of the supercapacitor participating in the real-time correction is: max f3=(R SC -H SC )-α3Flu3 H SC =h SC,P ·P SC +h SC,E ·E SC Among them, R SC represents the electricity price arbitrage profit of the supercapacitor in the real-time correction stage; H SC represents the rental cost of the supercapacitor; Flu3 represents the minute-level volatility of the power grid purchase curve, α3 represents the minute-level volatility weighting coefficient; t3 represents the index variable of the day-ahead stage, t3=1,2,…,15, Δt3 optimizes the decision time interval; p pv (t3) represents the peak-valley time-of-use electricity price at time t3; P SC,dis (t3) and P SC,ch (t3) respectively represent the discharge power and charging power of the supercapacitor at time t3; h SC,P and h SC,E They represent the unit power rental cost coefficient and unit capacity rental cost coefficient of the supercapacitor respectively; P SC and E SC Respectively represent the rated power and rated capacity of the supercapacitor; P real,grid (t3) represents the external power purchase amount at time t3, that is, the net load demand of the industrial park.

7. The shared hybrid energy storage control strategy considering multi-time scale coordination of source, load and storage according to claim 6 is characterized by: The constraint condition of the objective function f3 of the supercapacitor participating in the real-time correction is: (1) Correct power balance constraints in real time; P real,load (t3)-P real,grid (t3)-(P PS,G (t3)-P PS,P (t3))-(P BS,dis (t3)-P BS,ch (t3))=(P real,WP (t3)-P WP,ab (t3))+(P real,PV (t3)-P PV,ab (t3))+(P SC,dis (t3)-P SC,ch (t3)) Among them, P real,load (t3), P real,grid (t3), P real,WP (t3), P real,PV (t3) represent the forecast data of industrial park power load, external power purchase, wind power output and photovoltaic output at time t3 in the real-time correction optimization stage. (2) Physical constraints of supercapacitors; In the formula, SOC SC (t3) represents the charge state of the supercapacitor at time t3; η SC,ch and η SC,dis Respectively represent the charging efficiency and discharging efficiency of the supercapacitor; SOC SC,min and SOC SC,max Respectively represent the minimum and maximum value of the supercapacitor charge state; P SC,ch,max and η SC,dis,max Respectively represent the maximum charging power and maximum discharging power allowed by the supercapacitor; B SC,ch (t3) and B SC,dis (t3) are all Boolean variables, and state 1 represents the charging state and discharging state of the supercapacitor in the t3 period respectively.

Images