Grid-connected power generation power ultra-short-term prediction method based on optical storage optimization model

Abstract

The invention discloses a grid-connected generation power ultra-short-term prediction method based on a light storage optimization model. The method comprises the steps of establishing an energy storage distributed photovoltaic grid-connected generation power optimization model, providing constraint conditions, establishing a grid-connected generation power short-term prediction model and providing a rolling prediction mechanism. According to the invention, by considering the influence of energy storage charging and discharging behaviors on the grid-connected power generation power of light storage integration, the ultra-short-term prediction method for the distributed photovoltaic grid-connected power generation power containing energy storage, which comprehensively considers energy storage charging and discharging, photovoltaic power generation, on-grid and off-grid electricity prices and the like, is constructed.

Description

Technical Field

[0001] The invention relates to a photovoltaic power prediction method, and in particular to an ultra-short-term prediction method for grid-connected power based on a photovoltaic storage optimization model. Background Art

[0002] Large-scale grid-connected power generation of renewable energy represented by solar energy and wind energy has become an unstoppable development trend of new power systems. In particular, distributed photovoltaic power generation, as an important form of utilizing renewable energy, has been widely promoted and applied, and its profound impact on the power system is becoming more prominent. The academic and engineering communities of various countries have paid great attention to it, and a lot of research and application have been carried out in the field of photovoltaic power generation power prediction.

[0003] However, for distributed photovoltaic systems with energy storage, their impact on the power grid is determined by their grid-connected power generation rather than the power generation power of the photovoltaic system itself. Therefore, predicting their grid-connected power generation helps to accurately grasp the grid-connected characteristics of distributed photovoltaic systems with energy storage and their impact on the power grid, thereby achieving efficient absorption of distributed photovoltaic power generation.

[0004] Therefore, in order to solve the above problems, it is necessary to develop an ultra-short-term prediction method for grid-connected power generation based on the photovoltaic storage optimization model. Summary of the invention

[0005] The purpose of the present invention is to overcome the shortcomings of the prior art and to provide an ultra-short-term prediction method for grid-connected power generation based on a photovoltaic storage optimization model, taking into account the impact of energy storage charging and discharging behavior on the photovoltaic storage integrated grid-connected power generation power, and constructing an ultra-short-term prediction method for distributed photovoltaic grid-connected power generation with energy storage that comprehensively considers energy storage charging and discharging, photovoltaic power generation, and on-grid and off-grid electricity prices.

[0006] The object of the present invention is achieved as follows: a method for ultra-short-term prediction of grid-connected power generation based on a photovoltaic storage optimization model comprises the following steps:

[0007] S1. Establish a distributed photovoltaic grid-connected power generation optimization model with energy storage;

[0008] S11. Establish the total objective function of the power optimization model of distributed photovoltaic grid-connected power generation with energy storage; under the guidance of the electricity price mechanism, establish the total objective function of minimizing the total electricity cost of the next day, where the total electricity cost within a day includes the operating cost of the energy storage system and the purchase and sale of electricity costs at the gateway of the distributed photovoltaic system with energy storage;

[0009] S12, establishing sub-project objective functions; establishing objective functions for the operating costs of the energy storage system and the purchase and sale costs of electricity connected to the gateway respectively;

[0010] S2. Propose the constraint conditions of the power optimization model of distributed photovoltaic grid-connected power generation with energy storage, including: establishing energy storage operation constraints, grid-connected power purchase and sales constraints, and power balance constraints;

[0011] S3. Based on the distributed photovoltaic grid-connected power optimization model with energy storage and under constraints, a short-term prediction model for grid-connected power generation is established; specifically, based on the prediction of the distributed photovoltaic power generation power, according to the purchase and sale power price of the grid-connected power gateway, the above distributed photovoltaic grid-connected power optimization model with energy storage is used to obtain the optimal purchase and sale power of the grid-connected power gateway through optimization and solution, which is the short-term prediction result of the distributed photovoltaic grid-connected power generation with energy storage;

[0012] S4. Propose a rolling forecast mechanism; use the above-constructed short-term forecast model for grid-connected power generation to conduct rolling forecasts of grid-connected power generation from the current next time period to 24:00 at the end of the day according to a certain time period during the intraday operation.

[0013] Furthermore, the power optimization model of distributed photovoltaic grid-connected power generation with energy storage in step S1 is a mixed integer linear programming model, and the optimization solution in step S3 is solved by using the intlinprog function of the MATLAB optimization toolbox.

[0014] Furthermore, the total objective function for minimizing the total electricity cost of the next day in step S11 is as follows:

[0015]

[0016] Where T is the number of time periods in a day; C ES (t) is the operating cost of the energy storage system during period t (in yuan); C Grid (t) is the electricity purchase and sale cost of the connected gateway of the distributed photovoltaic system with energy storage. A positive value indicates electricity purchase, while a negative value indicates electricity sale (in yuan).

[0017] Furthermore, the operating cost objective function of the energy storage system in step S12 is as follows:

[0018]

[0019] Among them, N ES is the number of energy storage units, is the electricity cost of unit e when it is discharged, is the discharge power of unit e (in kW), δ t is the duration of a period;

[0020] Net discharge power of energy storage It can be expressed as (a positive value indicates discharge, and a negative value indicates charge):

[0021]

[0022] Among them, the energy storage charging power The electricity cost from other power sources has been included in the power generation cost of other power sources, and the energy storage charging cost is no longer included here.

[0023] Furthermore, the objective function of the electricity purchase and sale cost of the gateway in step S12 is as follows:

[0024]

[0025] Among them, N Grid The number of gateways for purchasing electricity from the external power grid. is the cost of purchasing electricity from the connecting gateway g (in yuan); is the income from selling electricity to the interconnection gateway g (in yuan);

[0026] in,

[0027]

[0028]

[0029] in, is the power purchased from the parallel gateway g (in kW); The price of electricity purchased; is the power sold to the parallel gateway g (in kW); The price of electricity sold;

[0030] The net power of the interconnection gateway g and the external power grid is (a positive value indicates power purchase, and a negative value indicates power sale):

[0031]

[0032] Furthermore, the energy storage operation constraints in step S2 include power upper and lower limit constraints, charge and discharge state constraints and SOC constraints;

[0033] The power upper and lower limit constraints are expressed as:

[0034]

[0035]

[0036] in, are the maximum discharge power and maximum charging power of unit e, respectively. are the minimum discharge power and minimum charging power of unit e (in kW), They are respectively the discharge state and the charge state. When the value is 1, it means discharge / charge, and when it is 0, it means no discharge / no charge;

[0037] Wherein, the charge and discharge state constraint is expressed as:

[0038] The charge and discharge states are mutually exclusive, that is, they satisfy the following constraints:

[0039]

[0040] At the same time, the following state transition constraints are met:

[0041]

[0042]

[0043] in, They are respectively the start-discharge state and the start-charge state. When the value is 1, it means that the discharge / charge is started at the current moment, and when it is 0, it means that there is no action at the current moment.

[0044] Wherein, the SOC constraint is expressed as:

[0045]

[0046]

[0047] S e (0) = S e (T) = S0

[0048] Among them, S e (t) is the SOC of unit e in period t, are the maximum and minimum allowable values ​​of SOC respectively, and S0 is the initial capacity of SOC.

[0049] Furthermore, the power purchase and sale power constraint of the gateway in step S2 is expressed as:

[0050]

[0051] in, They are the maximum power selling power and the maximum power purchasing power of the connected gateway g (in kW).

[0052] Furthermore, the power balance constraint in step S2 is expressed as:

[0053]

[0054] Among them, N PV is the number of photovoltaic units, Output power for photovoltaic unit p.

[0055] Furthermore, in step S4, a rolling forecast is performed on the grid-connected power generation power from the next full five minutes to the end of the day at 24:00, with a time period of 5 minutes.

[0056] Furthermore, in step S4, during rolling forecasting, it is necessary to roll-solve the constructed distributed photovoltaic grid-connected power generation optimization model with energy storage. In the rolling solution of the model, the rolling updated distributed photovoltaic self-generation power prediction result is used as the input of the distributed photovoltaic grid-connected power generation optimization model with energy storage.

[0057] Due to the adoption of the above technical scheme, the beneficial effects of the present invention are: by establishing a distributed photovoltaic grid-connected power generation optimization model with energy storage, and an ultra-short-term prediction method for grid-connected power generation based on the model, the problem of accurate prediction of distributed photovoltaic grid-connected power generation with energy storage can be effectively solved, and further analysis and application of the real grid-connected characteristics of distributed photovoltaics and their real impact on the power grid can be effectively supported. BRIEF DESCRIPTION OF THE DRAWINGS

[0058] Figure 1 It is a flow chart of the present invention. DETAILED DESCRIPTION

[0059] The technical solution of the present invention is further specifically described below through embodiments and in conjunction with the accompanying drawings.

[0060] like Figure 1 As shown, a method for ultra-short-term prediction of grid-connected power generation based on a photovoltaic storage optimization model includes the following steps:

[0061] S1. Establish a power optimization model for distributed photovoltaic grid-connected power generation with energy storage.

[0062] S11. Establish the total objective function of the power optimization model of distributed photovoltaic grid-connected power generation with energy storage; under the guidance of the electricity price mechanism, establish the total objective function of minimizing the total electricity cost of the next day, where the total electricity cost within a day includes the operating cost of the energy storage system and the purchase and sale of electricity costs at the gateway of the distributed photovoltaic system with energy storage. Specifically, the total objective function of minimizing the total electricity cost of the next day is as follows:

[0063]

[0064] Where T is the number of time periods in a day; C ES (t) is the operating cost of the energy storage system during period t (in yuan); C Grid (t) is the electricity purchase and sale cost of the connected gateway of the distributed photovoltaic system with energy storage. A positive value indicates electricity purchase, while a negative value indicates electricity sale (in yuan).

[0065] As a preferred embodiment, taking a 1MWp distributed photovoltaic system including 100kW / kWh energy storage as an example, with a time period resolution of 15 minutes, that is, T is 96, then the objective function for minimizing the total electricity cost of the next day is:

[0066]

[0067] Among them, C ES (t) and C Grid (t) Specifically as shown in step S12.

[0068] S12, establishing sub-project objective functions; establishing objective functions for the operating costs of the energy storage system and the electricity purchase and sales costs of the gateway.

[0069] Specifically, the operating cost objective function of the energy storage system is as follows:

[0070]

[0071] Among them, N ES is the number of energy storage units, is the electricity cost of unit e when it is discharged, is the discharge power of unit e (in kW), δ t is the duration of a period, preferably, δ t The time resolution can be 15 minutes. t When 15 minutes, or 0.25 hours, is used, T in the total objective function of minimizing the total electricity cost of the next day is 96.

[0072] Net discharge power of energy storage It can be expressed as (a positive value indicates discharge, and a negative value indicates charge):

[0073]

[0074] Among them, the energy storage charging power The electricity cost from other power sources has been included in the power generation cost of other power sources, and the energy storage charging cost is no longer included here.

[0075] As a preferred embodiment, the energy storage rated power in this embodiment is 100kW, and two 50kVA converters are used for grid connection, that is, the number of energy storage units is 2, and the cost of electricity when the unit e is discharged is It is 0.5 yuan / (kW·h).

[0076] Specifically, the objective function of the electricity purchase and sale cost of the gateway is as follows:

[0077]

[0078] Among them, NGrid The number of gateways for purchasing electricity from the external power grid. is the cost of purchasing electricity from the connecting gateway g (in yuan); is the income from selling electricity to the interconnection gateway g (in yuan);

[0079] in,

[0080]

[0081]

[0082] in, is the power purchased from the parallel gateway g (in kW); The price of electricity purchased; is the power sold to the parallel gateway g (in kW); The price of electricity sold;

[0083] The net power of the interconnection gateway g and the external power grid is (a positive value indicates power purchase, and a negative value indicates power sale):

[0084]

[0085] As a preferred embodiment, in this embodiment, the entire photovoltaic storage system is connected to the grid through a 1MVA transformer, that is, the number of connected gateways N Grid =1; power purchase price The electricity price is 0.3 yuan / (kW·h). It is 0.5 yuan / (kW·h).

[0086] S2. Propose the constraint conditions of the power optimization model of distributed photovoltaic grid-connected power generation with energy storage, including establishing energy storage operation constraints, grid-connected power purchase and sales constraints and power balance constraints.

[0087] The energy storage operation constraints include upper and lower power limit constraints, charge and discharge state constraints and SOC constraints.

[0088] Specifically, the power upper and lower limit constraints are expressed as:

[0089]

[0090]

[0091] in, are the maximum discharge power and maximum charging power of unit e, respectively. are the minimum discharge power and minimum charging power of unit e (in kW), They are discharge state and charge state respectively. When the value is 1, it means discharge / charge respectively, and when it is 0, it means no discharge / no charge respectively.

[0092] As a preferred embodiment, in this embodiment, the rated capacity of the energy storage unit is 50kVA, the maximum discharge power and the maximum charging power are 50kW, and the minimum discharge power of unit e is and minimum charging power (Unit: kW) are all 0.

[0093] when When is 1, the power upper and lower limit constraints can be expressed as:

[0094]

[0095]

[0096] when When it is 0, the upper and lower power limits are:

[0097]

[0098]

[0099] Right now is 0, which is consistent with It is 0, which means the expected power without discharging or charging.

[0100] Specifically, the charge and discharge state constraint is expressed as:

[0101] The charge and discharge states are mutually exclusive, that is, they satisfy the following constraints:

[0102]

[0103] At the same time, the following state transition constraints are met:

[0104]

[0105]

[0106] in, They are respectively the start-discharge state and the start-charge state. When the value is 1, it means that the discharge / charge is started at the current moment, and when it is 0, it means that there is no action at the current moment.

[0107] As a preferred embodiment, in this embodiment, at a certain moment is 0, that is, no discharge, and is 1, that is, the discharge state is started. If it is 0 and charging is not started, then at the next moment:

[0108]

[0109] That is, the next moment is the discharge state, which is consistent with the expected state of starting discharge at that moment.

[0110] Specifically, the SOC constraint is expressed as:

[0111]

[0112]

[0113] S e (0) = S e (T) = S0

[0114] Among them, S e (t) is the SOC of unit e in period t, are the maximum and minimum allowable values ​​of SOC respectively, and S0 is the initial capacity of SOC.

[0115] As a preferred embodiment, in this embodiment, the energy storage capacity of a single unit is 50kWh. In order to prevent overcharging, the maximum allowable SOC value can be 49kWh, and in order to prevent over-discharging, the minimum allowable SOC value can be 10kWh. S0 is the initial capacity of SOC, usually the minimum allowable SOC value, that is, 10kWh.

[0116] Specifically, the power purchase and sale power constraint of the parallel gateway is expressed as:

[0117]

[0118] in, They are the maximum power selling power and the maximum power purchasing power of the connected gateway g (in kW).

[0119] As a preferred embodiment, in this embodiment, the solar energy storage system is connected to the grid through a 1MVA transformer and the maximum power selling power of the connected gateway g is and maximum power purchase That is 1MW, or 1000kW, then the power purchase and sale power constraint of the gateway can be expressed as:

[0120]

[0121] Specifically, the power balance constraint is expressed as:

[0122]

[0123] Among them, N PV is the number of photovoltaic units, Output power for photovoltaic unit p.

[0124] As a preferred embodiment, in this embodiment, the total photovoltaic installed capacity is 1MWp, and a total of 40 inverters with a rated capacity of 40kVA are required for grid connection. Then the power balance constraint in this embodiment can be expressed as:

[0125]

[0126] S3. Based on the distributed photovoltaic grid-connected power optimization model with energy storage and under constraints, a short-term prediction model for grid-connected power generation is established;

[0127] Specifically, it includes, based on a given distributed photovoltaic power generation power prediction, according to the purchase and sale power price of the connected gateway, using the above-mentioned distributed photovoltaic grid-connected power generation optimization model with energy storage, and optimizing the solution. Specifically, the distributed photovoltaic grid-connected power generation optimization model with energy storage is a mixed integer linear programming model. When optimizing and solving, the intlinprog function of the MATLAB optimization toolbox can be used to solve it, and the optimal purchase and sale power of the connected gateway is obtained, which is the short-term prediction result of the distributed photovoltaic grid-connected power with energy storage.

[0128] In specific implementation, the output of photovoltaic units The optimization model established by using the distributed photovoltaic power generation prediction data, the objective function in step S1 and the constraint conditions in step S2 is the short-term prediction model of grid-connected power generation.

[0129] S4. Propose a rolling forecast mechanism; utilize the short-term forecast model for grid-connected power constructed as above, and during the daily operation, perform rolling forecast of the grid-connected power from the current next time period to 24:00 at the end of the day according to a certain time period; preferably, during the daily operation, perform rolling forecast of the grid-connected power from the current next full five minutes to 24:00 at the end of the day with a five-minute time period; specifically, during the rolling forecast, it is necessary to perform rolling solution of the constructed distributed photovoltaic grid-connected power optimization model with energy storage, and in the rolling solution of the model, use the rolling updated distributed photovoltaic self-generation power prediction result as the input of the distributed photovoltaic grid-connected power optimization model with energy storage.

[0130] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention rather than to limit it. Although the present invention has been described in detail with reference to the above embodiments, those skilled in the art should understand that the specific implementation modes of the present invention can still be modified or replaced by equivalents, and any modification or equivalent replacement that does not depart from the spirit and scope of the present invention should be included in the scope of the claims of the present invention.

Claims

1. A method for ultra-short-term prediction of grid-connected power generation based on a photovoltaic energy storage optimization model, characterized in that: The following steps are involved: S1. Establish a distributed photovoltaic grid-connected power generation optimization model with energy storage; S11. Establish the total objective function of the power optimization model of distributed photovoltaic grid-connected power generation with energy storage; under the guidance of the electricity price mechanism, establish the total objective function of minimizing the total electricity cost of the next day, where the total electricity cost within a day includes the operating cost of the energy storage system and the purchase and sale of electricity costs at the gateway of the distributed photovoltaic system with energy storage; S12, establishing sub-project objective functions; establishing objective functions for the operating costs of the energy storage system and the purchase and sale costs of electricity connected to the gateway respectively; S2. Propose the constraint conditions of the power optimization model of distributed photovoltaic grid-connected power generation with energy storage, including: establishing energy storage operation constraints, grid-connected power purchase and sales constraints, and power balance constraints; S3. Based on the distributed photovoltaic grid-connected power optimization model with energy storage and under constraints, a short-term prediction model for grid-connected power generation is established; Specifically, based on the prediction of the distributed photovoltaic power generation power, according to the purchase and sale power price of the grid-connected power grid, the above-mentioned distributed photovoltaic grid-connected power generation optimization model with energy storage is used to obtain the optimal grid-connected power purchase and sale power through optimization solution, which is the short-term prediction result of the distributed photovoltaic grid-connected power generation with energy storage; S4. Propose a rolling forecast mechanism; use the above-constructed short-term forecast model for grid-connected power generation to conduct rolling forecasts of grid-connected power generation from the current next time period to 24:00 at the end of the day according to a certain time period during the intraday operation.

2. According to claim 1, a method for ultra-short-term prediction of grid-connected power generation based on a photovoltaic energy storage optimization model is characterized by: The power optimization model of distributed photovoltaic grid-connected power generation with energy storage in step S1 is a mixed integer linear programming model, and the optimization solution in step S3 is solved by using the intlinprog function of the MATLAB optimization toolbox.

3. The method for ultra-short-term prediction of grid-connected power generation based on a photovoltaic energy storage optimization model according to claim 1 is characterized in that: The overall objective function for minimizing the total electricity cost of the next day in step S11 is as follows: Where T is the number of time periods in a day; C ES (t) is the operating cost of the energy storage system during period t (in yuan); C Grid (t) is the electricity purchase and sale cost of the connected gateway of the distributed photovoltaic system with energy storage. A positive value indicates electricity purchase, while a negative value indicates electricity sale (in yuan).

4. The method for ultra-short-term prediction of grid-connected power generation based on a photovoltaic energy storage optimization model according to claim 1 or 3, characterized in that: The operating cost objective function of the energy storage system in step S12 is as follows: Among them, N ES is the number of energy storage units, is the electricity cost of unit e when it is discharged, is the discharge power of unit e (in kW), δ t is the duration of a period; Net discharge power of energy storage It can be expressed as (a positive value indicates discharge, and a negative value indicates charge): Among them, the energy storage charging power The electricity cost from other power sources has been included in the power generation cost of other power sources, and the energy storage charging cost is no longer included here.

5. The method for ultra-short-term prediction of grid-connected power generation based on a photovoltaic energy storage optimization model according to claim 1 or 3, characterized in that: The objective function of the electricity purchase and sale cost of the gateway in step S12 is as follows: Among them, N Grid The number of gateways for purchasing electricity from the external power grid. is the cost of purchasing electricity from the connecting gateway g (in yuan); is the income from selling electricity to the interconnection gateway g (in yuan); in, in, is the power purchased from the parallel gateway g (in kW); The price of electricity purchased; is the power sold to the parallel gateway g (in kW); The price of electricity sold; The net power of the interconnection gateway g and the external power grid is (a positive value indicates power purchase, and a negative value indicates power sale):

6. The method for ultra-short-term prediction of grid-connected power generation based on a photovoltaic energy storage optimization model according to claim 1, characterized in that: The energy storage operation constraints in step S2 include power upper and lower limit constraints, charge and discharge state constraints and SOC constraints; The power upper and lower limit constraints are expressed as: in, are the maximum discharge power and maximum charging power of unit e, respectively. are the minimum discharge power and minimum charging power of unit e (in kW), They are respectively the discharge state and the charge state. When the value is 1, it means discharge / charge, and when it is 0, it means no discharge / no charge; Wherein, the charge and discharge state constraint is expressed as: The charge and discharge states are mutually exclusive, that is, they satisfy the following constraints: At the same time, the following state transition constraints are met: in, They are respectively the start-discharge state and the start-charge state. When the value is 1, it means that the discharge / charge is started at the current moment, and when it is 0, it means that there is no action at the current moment. Wherein, the SOC constraint is expressed as: S e (0)=S e (T)=S0 Among them, S e (t) is the SOC of unit e in period t, are the maximum and minimum allowable values ​​of SOC respectively, and S0 is the initial capacity of SOC.

7. The method for ultra-short-term prediction of grid-connected power generation based on a photovoltaic energy storage optimization model according to claim 1, characterized in that: The power consumption constraint of the gateway in step S2 is expressed as: in, They are the maximum power selling power and the maximum power purchasing power of the connected gateway g (in kW).

8. The method for ultra-short-term prediction of grid-connected power generation based on a photovoltaic energy storage optimization model according to claim 1, characterized in that: The power balance constraint in step S2 is expressed as: Among them, N PV is the number of photovoltaic units, Output power for photovoltaic unit p.

9. The method for ultra-short-term prediction of grid-connected power generation based on a photovoltaic energy storage optimization model according to claim 1, characterized in that: In step S4, a rolling forecast is made of the grid-connected power generation power from the next full five minutes to the end of the day at 24:00, with a five-minute time period.

10. The method for ultra-short-term prediction of grid-connected power generation based on a photovoltaic energy storage optimization model according to claim 1, characterized in that: In step S4, during rolling prediction, it is necessary to roll-solve the constructed distributed photovoltaic grid-connected power generation optimization model with energy storage. In the rolling solution of the model, the rolling updated distributed photovoltaic self-generation power prediction result is used as the input of the distributed photovoltaic grid-connected power generation optimization model with energy storage.

Images