Economic dispatching method for shared energy storage power station to participate in power grid auxiliary service

Abstract

The invention discloses an economic dispatching method for a shared energy storage power station to participate in power grid auxiliary service, which belongs to the field of renewable energy consumption, and comprises the following steps of: randomly sampling scene samples in each time period in a day by utilizing Latin hypercube sampling based on a renewable energy output probability model and an electric and thermal load probability model; a plurality of representative scenes are found out, the representative scenes are spliced to form a daily scene curve set, the daily scene curve set is substituted into a daily operation cost model, an operation strategy of the shared energy storage power station with the optimal cost is obtained, and the problems that output between wind power and photovoltaic power is influenced by climate, daily real-time scheduling of loads fluctuates and the like can be solved to a certain extent. The utilization rate of renewable energy sources is improved, and the power generation cost is reduced. The daily operation cost model considers cost factors in a plurality of aspects, such as power grid electricity purchase cost and service fee paid to the energy storage power station, can comprehensively evaluate economical efficiency and environmental influence of user energy consumption, and provides a scientific basis for energy management and decision making.

Description

Technical Field

[0001] The present invention belongs to the field of renewable energy consumption in power systems, and specifically relates to an economic dispatching method for a shared energy storage power station to participate in power grid auxiliary services. Background Art

[0002] In a virtual power plant, shared energy storage is an important component that can provide flexibility, stability and reliability, and provide added value and services to the power system and the market. The main reason is that shared energy storage power stations can be used for system power balancing, that is, releasing electricity when the power system demand peaks and storing electricity when it is low. This flexibility can help adjust the load curve of the power system and reduce the pressure and instability of the power system. It can also help solve the intermittent and unpredictable problems of renewable energy (such as wind and solar energy). It can store excess electricity generated by renewable energy and release it when needed, thereby improving the utilization and reliability of renewable energy. In summary, shared energy storage power stations in virtual power plants play an important role in improving the flexibility, reliability and efficiency of energy systems.

[0003] However, the output of renewable energy in shared energy storage is currently affected by climate, and daily load conditions will also vary with time periods. These unstable factors will affect the utilization rate of renewable energy and the cost of power generation, making it impossible to achieve the optimal economic scheduling of energy storage power stations. Summary of the invention

[0004] The purpose of the present invention is to provide an economic dispatching method for a shared energy storage power station to participate in grid auxiliary services, so as to solve the problems of low energy utilization and high dispatching cost caused by unstable output of renewable energy in the existing shared economic dispatching method.

[0005] In order to achieve the above object, the technical solution of the present invention is as follows: The present invention relates to an economic dispatching method for a shared energy storage power station to participate in power grid auxiliary services. The shared energy storage power station is a renewable energy power station, which comprises the following steps: S1. Construct the renewable energy output probability model and the electricity and heat load probability model; S2. Based on the renewable energy output probability model and the electric and thermal load probability model, Latin hypercube sampling is used to randomly sample scene samples at each time period of the day, and the scene reduction method is used to reduce the scene samples to find several representative scenes, and the representative scenes are spliced ​​to form a daily scene curve set; The scenario samples include renewable energy output samples and electricity and heat load samples. The daily scenario curve set formed by using the scenario reduction method and splicing includes renewable energy output curves and electricity and heat load curves. S3. Build a daily operation cost model, bring the daily scenario curve set into the daily operation cost model, and obtain the operation strategy of the shared energy storage power station with the best cost.

[0006] Preferably, the shared energy storage power station includes a wind power station and a photovoltaic power station, and the renewable energy output probability model constructed in S1 includes a wind turbine output probability model and a photovoltaic output probability model.

[0007] Preferably, the specific steps of constructing the wind turbine output probability model in S1 include: S1.1. Construct a wind speed probability model. The expression of the wind speed probability model is: (1), in, and Respectively in s The scale and shape parameters of the season t period, For s The actual wind speed during the season t period, is the probability density function of wind speed; (2), (3), (4), in, For s The mean wind speed in season t, For s The standard deviation of wind speed in season t, is the Gamma function; S1.2. Based on the wind speed probability model, the random wind speed size is obtained through inverse function transformation v ; S1.3. A wind turbine output probability model is constructed based on the wind speed. The expression of the wind turbine output probability model is: (5), in, P 1 is the fan output, v c , v f and v r are the cut-in wind speed, cut-out wind speed and rated wind speed of the wind turbine model respectively. P N is the rated output of the fan.

[0008] Preferably, the specific steps of constructing the photovoltaic output probability model in S1 include: S1.4. Construct a light intensity probability model. The expression of the light intensity probability model is: (6), in, for s The probability density function of the light intensity during the season t period is: for s season t The per unit value of light intensity during the period, and are the shape parameters of the Beta distribution, is the Gamma function; S1.5. Randomly obtain the light intensity per unit value of season t based on the light intensity probability model. S1.6. Based on s season t The photovoltaic output random model is constructed based on the per unit value of the light intensity in each period. The expression of the photovoltaic output random model is: (7), in, P 2 is the photovoltaic output size, is the rated capacity of the photovoltaic is the rated light intensity, is the reference value of light intensity.

[0009] Preferably, the probability density function of the electrical and thermal load probability model in S1 satisfies the following formula: (8), in, is the probability density function of the electrical and thermal load probability model, For the m The load in s season t The load size of the time period; , Respectively m The load in s season t Mean and standard deviation for the period.

[0010] Preferably, the specific steps of S2 randomly sampling scene samples at each time period in a day by using Latin hypercube sampling based on the renewable energy output probability model and the electric and thermal load probability model are as follows: S2.1. Calculate the cumulative probability distribution function based on the probability density function of the renewable energy output probability model and the probability model of the electricity and heat loads F ( P ), and divide the cumulative probability distribution function into Nt non-overlapping subintervals; the spacing between each interval is ;

[0011] S2.2. For any subinterval, generate a random number in the range [0,1] r t i , using random numbers r t i Calculate the cumulative probability function value corresponding to the interval, the calculation formula is: (9), in, q t i For the t period i The cumulative probability function value of subintervals is r t i is the generated random number; S2.3. The cumulative probability function value of the sub-interval q t i Substitute into the corresponding probability density function and calculate the sampling value of the sub-interval data by the inverse function , and then sample the scene of time period t , the calculation formula of the sampling value of the data is: (10), in, P t i For the t period i The sampled values ​​of the subintervals, After the above steps, the sampling is obtained The scene of the period is .

[0012] Preferably, the specific steps of S2 using a scene reduction method to reduce scene samples include: S2.4. Delete the scene generated by the above steps The number of scenes with sampling values ​​less than 0 in the reduction is N t ', set the target scene number , reduce scene collection and preserve scene collections ; S2.5. Traverse the scene, find another scene that is closest to it, calculate the probability product of the scene and the scene that is closest to it, and find the corresponding scene with the smallest probability product, and record the scene number of the scene; The distance calculation formula between two scenes is: (12), in, The time period number is t i and j The distance between the two scenes, is the scene numbered i in period t, is the scene numbered j in period t; The probability product z t i The calculation formula is: (13), Recorded scene number l * means: (14), S2.6. Update the reduction scene set separately and preserve scene collections , determine whether the number of scenes in the retained scene set reaches the target number of scenes, if so, complete the single-period scene reduction and output several representative scene sets , otherwise return to S2.5.

[0013] Preferably, in S2, the specific steps of splicing the representative scenes to form a daily scene curve set are: S2.7. Compare the representative scene set according to the specific sampling data size Sort and get the sorted representative scene set , set the initial period t to 1, set the scene fluctuation threshold ε and the initial scene ; S2.8. For the initial , define other scenarios Any scene in ,match The corresponding next time period scene ,in For collection The random number in ; S2.9. Determine whether the period is the last period. If not, t Update to the next period and return to S2.8. If yes, The corresponding scenes are spliced ​​together to obtain a set of daily scene curves.

[0014] Preferably, the cost in the daily operation cost model constructed in S3 includes the power purchase fee of the power grid and the cost of using the energy storage power station. The expression of the power purchase fee of the power grid is: (15), in, C 1 i The cost of purchasing electricity from the power grid, T is the number of scheduling cycle time periods, is the number of target scenes, γ is the price coefficient of electricity purchased from the power grid, α ( i ) is the scene i The probability of occurrence, For the scene i The power purchased from the grid during the next period t; The expression of the cost of using the energy storage power station is: (16), in, C 2 i The cost of using the energy storage power station. δ ( i ) is the scene i The service fee coefficient paid to the energy storage power station is , Respectively for scenes i The next t period uses the charging and discharging power of the energy storage power station.

[0015] Preferably, the S3 daily operation cost model also satisfies the following constraints: The electric power balance constraint is expressed as: (17) in, For the scene i Down t PV power generation during the period; For the scene i The wind power generation power in the next period t; For the scene i Down t The electrical load power of the time period; The charging and discharging power constraints of users using shared energy storage power stations are expressed as: (18), in, and Respectively for scenes i The charging power and discharging power of the shared energy storage power station are as follows: The maximum power of charging and discharging using the shared energy storage station, and Respectively for scenes i The discharge status bit and charge status bit of the shared energy storage power station are used; The continuity constraint of the state of charge of the energy storage power station is expressed as: (19), (20), in, E t , E t-1 They are t Moment and t -1 The capacity of the energy storage power station at the moment, and They are t The charging and discharging power of the energy storage station itself at all times, and are the charging and discharging efficiency of the energy storage power station, and are the maximum and minimum values ​​of the energy storage power station under charge state, μ is the charge and discharge coefficient of energy storage; The charging and discharging power constraints of the energy storage power station itself are expressed as: (twenty one), in, and are the charging power and discharging power of the energy storage power station respectively. and They are respectively the charging status and discharging status of the energy storage power station. The maximum value of charging and discharging for the energy storage power station; The charging and discharging power balance constraint of the energy storage power station is expressed as: (twenty two).

[0016] Compared with the prior art, the technical solution provided by the present invention has the following beneficial effects: 1. The energy storage economic dispatch method for the shared energy storage power station participating in the grid auxiliary service involved in the present invention is based on the renewable energy output probability model and the electric and thermal load probability model. It uses Latin hypercube sampling to randomly sample scene samples in each time period of a day, find out several representative scenes, splice the representative scenes to form a daily scene curve set, and bring it into the daily operation cost model to obtain the cost-optimal operation strategy of the shared energy storage power station. It can solve the problems of the output between wind power and photovoltaic power being affected by the climate and the daily real-time dispatch fluctuations of load to a certain extent, thereby improving the utilization rate of renewable energy and reducing the cost of power generation.

[0017] 2. The economic dispatch method for the shared energy storage power station participating in the grid auxiliary service involved in the present invention uses the Latin hypercube sampling method for random sampling, and then adopts the scenario reduction method to reduce the scenario samples to find out several representative scenarios, thereby reducing the time consumption of subsequent scenario splicing. By splicing representative scenarios, the continuity of the scenarios, that is, the continuity of the user load level, is maintained, while taking into account the volatility of the residents' load, so that the generated scenarios are as consistent as possible with the residents' electricity consumption habits, thereby achieving the optimal economic dispatch of the energy storage power station.

[0018] 3. The daily operating cost model constructed by the economic dispatch method for the shared energy storage power station participating in the grid auxiliary service involved in the present invention includes the power purchase cost of the power grid and the cost of using the energy storage power station. By comprehensively considering these cost factors, the economic efficiency and environmental impact of energy consumption are comprehensively evaluated, providing a scientific basis for energy management and decision-making. BRIEF DESCRIPTION OF THE DRAWINGS

[0019] Figure 1 It is a flow chart of the economic dispatch method of the shared energy storage power station participating in the grid ancillary services; Figure 2 It is a graph of the relationship between wind turbine output and wind speed; Figure 3 It is a graph of the relationship between photovoltaic output and light intensity; Figure 4 This is the output diagram of photovoltaic and wind power after reduction; Figure 5 is the load size of two users 1 and 2 after the reduction; Figure 6 is the electric load balance curve of user 1; Figure 7 is the electric load balance curve of user 2; Figure 8 It is the charging and discharging power and power status curve of the shared power station. DETAILED DESCRIPTION

[0020] In order to further understand the content of the present invention, the present invention is described in detail in conjunction with examples. The following examples are used to illustrate the present invention but are not used to limit the scope of the present invention.

[0021] Refer to the attached Figure 1 As shown, the present invention relates to an economic dispatching method for a shared energy storage power station to participate in power grid auxiliary services, wherein the shared energy storage power station is a renewable energy power station, and the renewable energy output probability model includes a wind turbine output probability model and a photovoltaic output probability model. The energy storage economic dispatching method includes the following steps: S1. Construct the renewable energy output probability model and the electricity and heat load probability model; The probability model of wind turbine output depends on the probability model of wind speed and the relationship between wind turbine output and wind speed. It is usually assumed that wind speed follows Weibull distribution, such as Figure 2 As shown in Figure 2, the specific steps of constructing the wind turbine output probability model include: S1.1. Construct a wind speed probability model. The expression of the wind speed probability model is: (1), in, and Respectively in s The scale and shape parameters of the season t period, For s The actual wind speed during the season t period, is the probability density function of wind speed; (2), (3), (4), in, For s The mean wind speed in season t, For s The standard deviation of wind speed in season t, is the Gamma function; S1.2. Based on historical typical data, the corresponding scale parameters and shape parameters can be obtained, which are introduced into the wind speed probability model and the random wind speed size can be obtained through inverse function transformation. v ; S1.3. As an energy conversion system, after the input wind speed is determined, the output power of the wind turbine is mainly determined by its own parameters, that is, the expression of the wind turbine output probability model is: (5), in, P 1 is the fan output, v c ,v f and v r are the cut-in wind speed, cut-out wind speed and rated wind speed of the wind turbine model respectively. P N is the rated output of the fan.

[0022] Photovoltaic power generation is a renewable energy generation technology that uses the most abundant energy to create the cleanest electricity. It belongs to the energy supply part of distributed energy stations and is particularly important for improving the cleanliness and reliability of energy. The output power of photovoltaic power generation depends on the degree of radiation received per unit area of ​​the photovoltaic panel. According to statistics, the probability model of light intensity over a period of time can be approximately regarded as obeying the Beta distribution, such as Figure 3 shown.

[0023] The specific steps of constructing the photovoltaic output probability model include: S1.4. Construct a light intensity probability model. The expression of the light intensity probability model is: (6), in, for s season t The probability density function of the light intensity during the period, for s season t The per unit value of light intensity during the period, and are the shape parameters of the Beta distribution, is the Gamma function; S1.5. Similarly, by changing the inverse function, we can obtain the random light intensity that obeys the beta distribution, and then randomly obtain based on the light intensity probability model s The per unit value of light intensity in season t, S1.6. Based on s season t The photovoltaic output random model is constructed based on the per unit value of the light intensity in each period. In the photovoltaic output random model, the expression of photovoltaic output is: (7), in, P 2 is the photovoltaic output size, is the rated capacity of the photovoltaic is the rated light intensity, is the reference value of light intensity.

[0024] This method assumes that the electrical and thermal loads follow a normal distribution. Therefore, the probability density function of the electrical and thermal load probability model satisfies the following formula: (8), in, is the probability density function of the electrical and thermal load probability model, For the m The load in s season t The load size of the time period; , Respectively m The load in s Mean and standard deviation for the seasonal t period.

[0025] S2. Based on the renewable energy output probability model and the electric and thermal load probability model, Latin hypercube sampling is used to randomly sample scene samples at each time period of the day, and the scene reduction method is used to reduce the scene samples to find several representative scenes, and the representative scenes are spliced ​​to form a daily scene curve set; The scenario samples include renewable energy output samples and electricity and heat load samples. The daily scenario curve set formed by using the scenario reduction method and splicing includes renewable energy output curves and electricity and heat load curves.

[0026] S2 first uses Latin hypercube sampling to generate a large number of random scenes to ensure that all sampling points are sampled. The specific steps are as follows: S2.1. Calculate the cumulative probability distribution function based on the probability density function of the renewable energy output probability model and the probability model of the electricity and heat loads F ( P ), where, for the wind turbine output probability model, the cumulative probability distribution function F ( P ) corresponds to the wind turbine output; for the photovoltaic output probability model, the cumulative probability distribution function F ( P ) corresponds to the photovoltaic output. For the probability model of electricity and heat load, the cumulative probability distribution function F ( P ) corresponds to the load size, and the cumulative probability distribution function is divided into N t non-overlapping subintervals, each with a spacing of 1 / N t ; S2.2. For N t Generate a random number in the range [0,1] from any subinterval of r t i , use random numbers to calculate the cumulative probability function value of the corresponding subinterval, the calculation formula is: (9), in, q t i for t Time period i The cumulative probability function value of subintervals is r t i is the generated random number; S2.3. The cumulative probability function value of the sub-interval q t i Substitute into the corresponding probability density function and calculate the sampling value of the sub-interval data by the inverse function , and then sample the scene of time period t , the calculation formula of the sampling value of the data is: (10), in, P t i is the sampling value of the i-th subinterval in time period t.

[0027] To ensure the accuracy of the simulation, a large number of scenes usually need to be generated in each period, but the subsequent splicing of each scene takes a long time. Therefore, in order to comprehensively consider the computational efficiency and accuracy, it is necessary to screen the scenes t generated in each period and retain a representative part of the scenes. Therefore, this step also uses the scene reduction method to reduce the scene samples. The specific steps include:

[0028] S2.4. Delete the scene generated by the above steps The number of scenes with sampling values ​​less than 0 is reduced to N t ', set the target scene number , reduce scene collection and preserve scene collections ; (11), S2.5. Traverse the scene, find another scene that is closest to it, calculate the probability product of the scene and the scene that is closest to it, and find the corresponding scene with the smallest probability product, and record the scene number of the scene; The distance calculation formula between two scenes is: (12), in, for t The distance between two scenes with time interval numbers i and j. The scene number is equal to the sub-interval number. is the scene numbered i in period t, is the scene numbered j in period t; The probability product z t i The calculation formula is: (13), Recorded scene number l * means: (14), S2.6. Update the reduction scene set separately and preserve scene collections , determine whether the number of scenes in the retained scene set reaches the target number of scenes, if so, complete the single-period scene reduction and output several representative scene sets , otherwise return to S2.5.

[0029] After obtaining the representative scenes of each time period, these single-time period scenes need to be connected. Although the scenes of each time period have been reduced, the number of scene sets obtained by directly connecting all single-time period scenes is still very large. For example, assuming that 2 representative scenes are generated for each time period, if a day is divided into 24 time periods, these single scenes will be directly connected to generate a total of 2 24 In order to generate load scenarios more efficiently, this step combines representative scenarios to form a daily scenario curve set. The specific steps are as follows:

[0030] S2.7. Compare the representative scene set according to the specific sampled data size (wind turbine output size, photovoltaic output size and load size). Sort and get the sorted representative scene set , set the initial period t is 1, setting the scene fluctuation threshold ε and the initial scene ; S2.8. For the initial , define other scenarios Any scene in ,match The corresponding next time period scene ,in For collection The random number in ; S2.9. Determine whether this period is the last period. If not, update period t to the next period and return to S2.8. If so, The corresponding scenes are spliced ​​together to obtain a set of daily scene curves.

[0031] S3. Construct a daily operation cost model, bring the daily scenario curve set into the daily operation cost model, and obtain the operation strategy of the shared energy storage power station with the best cost; the cost in the daily operation cost model includes the power purchase fee of the power grid and the cost of using the energy storage power station. The expression of the power purchase fee of the power grid is: (15), in, C 1 i The cost of purchasing electricity from the power grid, T is the number of scheduling cycle time periods, is the number of target scenes, γ is the price coefficient of electricity purchased from the power grid, α ( i ) is the scene i The probability of occurrence, For the scene i Down t The power purchased from the grid during the period; The expression of the cost of using the energy storage power station is: (16), in, C 2 i The cost of using the energy storage power station. δ ( i ) is the scene i The service fee coefficient paid to the energy storage power station is , Respectively for scenes i The next t period uses the charging and discharging power of the energy storage power station.

[0032] The above-mentioned daily operation cost model also satisfies the following constraints: The electric power balance constraint is expressed as: (17), in, For the scene i The photovoltaic power generation in the next period t; For the scene i The wind power generation power in the next period t; For the scene i The electric load power in the next period t; The charging and discharging power constraints of users using shared energy storage power stations are expressed as: (18), in, and Respectively for scenes iThe charging power and discharging power of the shared energy storage power station are as follows: The maximum power of charging and discharging using the shared energy storage station, and Respectively for scenes i The discharge status bit and charge status bit of the shared energy storage power station are used; The continuity constraint of the state of charge of the energy storage power station is expressed as: (19), (20), in, E t , E t-1 They are t Moment and t -1 The capacity of the energy storage power station at the moment, and They are t The charging and discharging power of the energy storage station itself at all times, and are the charging and discharging efficiency of the energy storage power station, and are the maximum and minimum values ​​of the energy storage power station under charge state, μ is the charge and discharge coefficient of energy storage; The charging and discharging power constraints of the energy storage power station itself are expressed as: (twenty one), in, and are the charging power and discharging power of the energy storage power station respectively. and They are respectively the charging status and discharging status of the energy storage power station. The maximum value of charging and discharging for the energy storage power station; The charging and discharging power balance constraint of the energy storage power station is expressed as: (twenty two).

[0033] Effect example: Set the parameters, set the initial state of the shared energy storage power station to 0.2, and the maximum and minimum charge states to 0.9 and 0.1 respectively. The service fee of the shared service power station is 0.28 yuan / kW·h. The maximum power of each user charging and discharging using the shared energy storage power station is 1000kW. The capacity cost of the energy storage power station uses the winning bid price of the battery of a certain energy storage project in 2022 of 1050 yuan / (kW·h), the power cost uses the winning bid price of the project of 1000 / (kW·h), and the operation and maintenance cost is 70 yuan / (a·kW). The power purchase price of the power grid is shown in Table 1:

[0034] Table 1 Electricity price parameters category Time Price of electricity purchased from the power grid (yuan / kW·h) peak 08：00~12:00、17:00~21：00 1.40 flat 12：00~17:00、21：00~24：00 0.80 valley 0：00~8：00 0.37 Simulation results description: Figure 4 , Figure 5 Latin hypercube sampling is used to obtain typical wind and light output power curves and load power curves of users 1 and 2. The length of each scheduling period is 15 minutes. The configuration result of the shared energy storage power station is 2872kW·h, the maximum charge and discharge power is 635kW, and the charge and discharge power and power status of the energy storage power station are as follows Figure 8 Among them, if the charging and discharging power value of the energy storage station is negative, it means that the energy storage station is charging; if the charging and discharging power value is positive, it means that the energy storage station is discharging.

[0035] Figure 6 It can be seen that during the period of 00:00-09:00, the photovoltaic output of user 1 is less than the user's load demand. User 1 replenishes electricity by discharging from the shared energy storage power station and purchasing electricity from the power grid. During the period of 09:00-16:00, the photovoltaic output is greater than the load demand of user 1. The remaining electricity in user 1 is stored in the shared energy storage power station to avoid power abandonment. During the period of 13:00-14:00, the photovoltaic output reaches a maximum value of 750kW, and user 1 uses the shared energy storage power station to charge to a maximum power of 590kW. During the period of 16:00-24: 00, the photovoltaic output is less than the user's load demand. Since this period belongs to the peak period and the flat period of the power price of the power grid, in order to minimize the operating cost, user 1 did not purchase electricity from the power grid, and used the shared energy storage power station to discharge to meet the load demand. During the period of 16:30-16:45, the power discharged by the user using the energy storage power station reached a maximum of 175kW.

[0036] Figure 7 It can be seen that the load and wind power output of user 2 are highly volatile. In the period of 00:30-01:30, user 2 purchased electricity from the power grid, with a maximum power purchase of 25kW. Others use the shared energy storage power station to charge and consume the remaining wind power. When the wind power output is less than the power load of user 2, the user uses the shared energy storage power station to discharge to meet the load demand. In the period of 16:30-16:30, the power discharged by user 2 using the energy storage power station reached a maximum of 170kW; in the period of 04:30-05: 300, the power charged by user 2 using the energy storage power station reached a maximum of 115kW.

[0037] Depend on Figure 6 and Figure 7 It can be seen that the power loads of users 1 and 2 have reached a balanced state, and there is no abandonment of solar power, wind power, or electricity. At the same time, the shared energy storage power station returns to its initial operating state after a cycle of operation, ensuring the normal operation of the energy storage power station in the next cycle.

[0038] In order to analyze the economic feasibility of introducing a shared energy storage power station, the following two scenarios are set for comparative analysis: Scenario 1: The user does not configure energy storage.

[0039] Scenario 2: Users access a shared energy storage power station.

[0040] Table 2 Optimization scheduling results Is the energy storage power station connected? Abandoned electricity (kw·h) Electricity purchased from the power grid / (kw·h) Service fee / (yuan) Operating cost / (yuan) Access 0 207.8 1160.5 1190.8 No access 2108 1570 0 1320.7 100% of the power generated by renewable energy has been absorbed, which means that more renewable energy can be incorporated into the grid and utilized; the total operating cost of users is reduced, because the operation of shared energy storage power stations can flexibly adjust the supply of electricity, so that users do not have to rely on traditional power grids all the time, thereby reducing electricity bills; the power purchased from the grid has decreased, which also reflects that users are less dependent on traditional power grids and rely more on their own renewable energy and shared energy storage power stations; the complementarity of user electricity consumption behavior makes the overall electricity load more stable, which contributes to the stable operation of the power grid; due to the complementarity of user electricity consumption behavior, the scale of energy storage required by users is greatly reduced, which means that shared energy storage power stations can more effectively meet user needs and reduce system construction and operation costs.

[0041] The present invention is described in detail above in conjunction with the embodiments, but the contents described are only preferred embodiments of the present invention and cannot be considered to limit the scope of implementation of the present invention. All equivalent changes and improvements made according to the scope of application of the present invention should still fall within the scope of the patent coverage of the present invention.

Claims

1. An economic dispatch method for a shared energy storage power station to participate in grid auxiliary services, wherein the shared energy storage power station is a renewable energy power station, characterized in that: It includes the following steps: S1. Construct the renewable energy output probability model and the electricity and heat load probability model; S2. Based on the renewable energy output probability model and the electric and thermal load probability model, Latin hypercube sampling is used to randomly sample scene samples at each time period of the day, and the scene reduction method is used to reduce the scene samples to find several representative scenes, and the representative scenes are spliced ​​to form a daily scene curve set; The scenario samples include renewable energy output samples and electricity and heat load samples. The daily scenario curve set formed by using the scenario reduction method and splicing includes renewable energy output curves and electricity and heat load curves. S3. Build a daily operation cost model, bring the daily scenario curve set into the daily operation cost model, and obtain the operation strategy of the shared energy storage power station with the best cost.

2. The economic dispatch method for a shared energy storage power station to participate in grid auxiliary services according to claim 1, characterized in that: The shared energy storage power station includes a wind power station and a photovoltaic power station, and the renewable energy output probability model constructed in S1 includes a wind turbine output probability model and a photovoltaic output probability model.

3. The economic dispatching method for a shared energy storage power station to participate in grid auxiliary services according to claim 2, characterized in that: The specific steps of constructing the wind turbine output probability model in S1 include: S1.

1. Construct a wind speed probability model. The expression of the wind speed probability model is: （1）， in, and Respectively in s season t The scale and shape parameters of the time period, For s The actual wind speed during the season t period, is the probability density function of wind speed; （2）， （3）， （4）， in, For s The mean wind speed in season t, For s The standard deviation of wind speed in season t, is the Gamma function; S1.

2. Based on the wind speed probability model, the random wind speed size is obtained through inverse function transformation v ; S1.

3. A wind turbine output probability model is constructed based on the wind speed. The expression of the wind turbine output probability model is: （5）， in, P 1 is the fan output, v c , v f and v r are the cut-in wind speed, cut-out wind speed and rated wind speed of the wind turbine model respectively. P N is the rated output of the fan.

4. The economic dispatching method for a shared energy storage power station to participate in grid auxiliary services according to claim 2, characterized in that: The specific steps of constructing the photovoltaic output probability model in S1 include: S1.

4. Construct a light intensity probability model. The expression of the light intensity probability model is: （6）， in, for s The probability density function of the light intensity during the season t period is: for s The per unit value of light intensity during the season t period, and are the shape parameters of the Beta distribution, is the Gamma function; S1.

5. Randomly obtain based on the light intensity probability model s The per unit value of light intensity in season t, S1.

6. Based on s season t The photovoltaic output random model is constructed based on the per unit value of the light intensity in each period. The expression of the photovoltaic output random model is: （7）， in, P 2 is the photovoltaic output size, is the rated capacity of the photovoltaic is the rated light intensity, is the reference value of light intensity.

5. The economic dispatch method for a shared energy storage power station to participate in grid auxiliary services according to claim 1, characterized in that: The probability density function of the electrical and thermal load probability model in S1 satisfies the following formula: （8）， in, is the probability density function of the electrical and thermal load probability model, For the m The load in s season t The load size of the time period; , Respectively m The load in s season t Mean and standard deviation for the period.

6. The economic dispatch method for a shared energy storage power station to participate in grid auxiliary services according to claim 1, characterized in that: The specific steps of S2 for randomly sampling scene samples at each time period in a day using Latin hypercube sampling based on the renewable energy output probability model and the electric and thermal load probability model are as follows: S2.

1. Calculate the cumulative probability distribution function based on the probability density function of the renewable energy output probability model and the probability model of the electricity and heat loads F ( P ), and divide the cumulative probability distribution function into N t non-overlapping subintervals; the spacing between each interval is ; S2.

2. For any subinterval, generate a random number in the range [0,1] r t i , using random numbers r t i Calculate the cumulative probability function value corresponding to the interval, and the calculation formula is: （9）， in, q t i For the t period i The cumulative probability function value of subintervals is r t i is the generated random number; S2.

3. The cumulative probability function value of the sub-interval q t i Substitute into the corresponding probability density function and calculate the sampling value of the sub-interval data by the inverse function , and then sample the scene of time period t , the calculation formula of the sampling value of the data is: （10）， in, P t i is the sampling value of the ith subinterval in period t, After the above steps, the sampling is obtained The scene of the period is .

7. The economic dispatch method for a shared energy storage power station to participate in grid auxiliary services according to claim 6, characterized in that: The specific steps of S2 using the scene reduction method to reduce the scene samples include: S2.

4. Delete the scene generated by the above steps The number of scenes with sampling values ​​less than 0 is reduced to N t ', set the target scene number , reduce scene collection and preserve scene collections ; S2.

5. Traverse the scene, find another scene that is closest to it, calculate the probability product of the scene and the scene that is closest to it, and find the corresponding scene with the smallest probability product, and record the scene number of the scene; The distance calculation formula between two scenes is: （12）， in, is the distance between the two scenes numbered i and j in period t, is the scene numbered i in period t, is the scene numbered j in period t; The probability product z t i The calculation formula is: （13）， Recorded scene number l * means: （14）， S2.

6. Update the reduction scene set separately and preserve scene collections , determine whether the number of scenes in the retained scene set reaches the target number of scenes. If so, complete the single-period scene reduction and output several representative scene sets , otherwise return to S2.

5.

8. The economic dispatch method for a shared energy storage power station to participate in grid auxiliary services according to claim 7, characterized in that: In S2, the specific steps of splicing the representative scenes to form a daily scene curve set are: S2.

7. Compare the representative scene set according to the specific sampling data size Sort and get the sorted representative scene set , set the initial period t is 1, setting the scene fluctuation threshold ε and the initial scene ; S2.

8. For the initial , define other scenarios Any scene in ,match The corresponding next time period scene ,in For collection The random number in ; S2.

9. Determine whether this period is the last period. If not, update period t to the next period and return to S2.

8. If so, The corresponding scenes are spliced ​​together to obtain a set of daily scene curves.

9. The economic dispatch method for a shared energy storage power station to participate in grid auxiliary services according to claim 1, characterized in that: The cost in the daily operation cost model constructed in S3 includes the power purchase fee of the power grid and the cost of using the energy storage power station. The expression of the power purchase fee of the power grid is: （15）， in, C 1 i The cost of purchasing electricity from the power grid, T is the number of scheduling cycle time periods, is the number of target scenes, γ is the price coefficient of electricity purchased from the power grid, α ( i ) is the scene i The probability of occurrence, For the scene i The power purchased from the grid during the next period t; The expression of the cost of using the energy storage power station is: （16）， in, C 2 i The cost of using the energy storage power station. δ ( i ) is the scene i The service fee coefficient paid to the energy storage power station is , They are respectively the charging and discharging power of the energy storage power station used in period t under scenario i.

10. The economic dispatch method for a shared energy storage power station to participate in grid auxiliary services according to claim 9, characterized in that: The S3 daily operation cost model also satisfies the following constraints: The electric power balance constraint is expressed as: （17） in, is the photovoltaic power generation power in period t under scenario i; is the wind power generation power in period t under scenario i; is the electric load power in period t under scenario i; The charging and discharging power constraints of users using shared energy storage power stations are expressed as: （18）， in, and Respectively for scenes i The charging power and discharging power of the shared energy storage power station are as follows: The maximum power of charging and discharging using the shared energy storage station, and Respectively for scenes i The discharge status bit and charge status bit of the shared energy storage power station are used; The charge state continuity constraint of the energy storage power station is expressed as: （19）， （20）， in, E t , E t-1 They are t Moment and t -1 The capacity of the energy storage power station at the moment, and They are t The charging and discharging power of the energy storage station itself at all times, and are the charging and discharging efficiency of the energy storage power station, and are the maximum and minimum values ​​of the energy storage power station under charge state, μ is the charge and discharge coefficient of energy storage; The charging and discharging power constraints of the energy storage power station itself are expressed as: （21）， in, and are the charging power and discharging power of the energy storage power station respectively. and They are respectively the charging status and discharging status of the energy storage power station. The maximum value of charging and discharging for the energy storage power station; The charging and discharging power balance constraint of the energy storage power station is expressed as: （22）。

Images