System and method for generating control strategy of energy storage system based on dynamic electricity price

Abstract

The invention provides a control strategy generation system and method of an energy storage system based on dynamic electricity price. The control strategy generation system comprises a prediction module, a strategy generation module and a cloud platform. The prediction module comprises a historical data acquisition module, a first prediction module and a second prediction module; the historical data acquisition module is respectively connected with the first prediction module and the second prediction module; the first prediction module is also connected with the strategy generation module; the second prediction module is also connected with the strategy generation module; the strategy generation module is also connected with the cloud platform; and the cloud platform is also connected with the historical data acquisition module. The method effectively adapts to energy planning under a dynamic electricity price background, and maximizes economic benefits of users and saves electric charges by optimizing peak clipping and valley filling operations.

Description

Technical Field

[0001] The present application relates to an energy storage system, and in particular to the technical field of a control strategy for an energy storage system based on dynamic electricity prices. Background Art

[0002] In modern energy management, with the widespread application of renewable energy and the promotion of dynamic electricity price mechanisms, how to efficiently manage home energy storage systems to reduce electricity costs has become a key challenge. Existing energy storage systems often face problems such as inaccurate photovoltaic power generation forecasts, unstable load demand, and dynamic electricity price fluctuations. In order to optimize the operation of energy storage systems, current technologies need to improve forecast accuracy, optimize strategy generation, and flexibly respond to policy changes.

[0003] In view of this, this application is proposed. Summary of the invention

[0004] The present invention aims to solve these problems and provides a control strategy generation system and a control strategy generation method for an energy storage system based on dynamic electricity prices.

[0005] On the one hand, the present application provides a control strategy generation system for an energy storage system based on dynamic electricity prices, comprising: a prediction module, a strategy generation module and a cloud platform;

[0006] The prediction module includes a historical data acquisition module, a first prediction module and a second prediction module; the historical data acquisition module is connected to the first prediction module and the second prediction module respectively; the first prediction module is also connected to the strategy generation module; the second prediction module is also connected to the strategy generation module; the strategy generation module is also connected to the cloud platform; the cloud platform is also connected to the historical data acquisition module.

[0007] Furthermore, the historical data acquisition module is used to obtain historical operation data and current operation data from the cloud platform; the historical operation data includes historical photovoltaic power generation power P hist and historical load power L hist The current operating data includes the current photovoltaic power generation power P now and the current load power L now ;

[0008] The first prediction module is used to predict the future photovoltaic power generation power P in a certain time period in the future based on the historical operation data and the current operation data. fcst ;

[0009] The second prediction module is used to predict the future load power L in a certain time period in the future based on the historical operation data and the current operation data. fcst ;

[0010] The strategy generation module is used to generate a strategy based on the future photovoltaic power generation power P fcst and future load power L fcst , using the output of the linear programming model to select the appropriate operation mode and generate control instructions;

[0011] The cloud platform is used to obtain the current operating data of the energy storage system through the collector, store and update the historical data, and issue the control instructions to each energy storage system.

[0012] Furthermore, the output of the strategy generation module includes P buy , P sell , P discharge and P charge , the linear programming model is:

[0013]

[0014] Where: T is the number of time periods for prediction, with H minutes as one time period, H is an integer, M sell is the electricity price for each time period, P sell is the electricity sold in each time period, M buy P is the electricity purchase price for each time period. buy The power purchased in each time period, Maximize PROFIT is the maximum profit.

[0015] On the other hand, the present application also provides a method for generating a control strategy for an energy storage system based on dynamic electricity prices using the above system, comprising the following steps:

[0016] Step 1: Acquire the historical operation data and the current operation data of the energy storage system from the cloud platform;

[0017] Step 2: Calculate the future photovoltaic power generation power P based on the historical operation data and the current operation data fcst According to the historical operation data and the current operation data, the future load power L is predicted. fcst Make predictions;

[0018] Step 3: Based on the future photovoltaic power generation power P fcst 、Future load power L fcst and the output of the linear programming model, selecting a suitable operation mode for the energy storage system and generating a control instruction;

[0019] Step 4: The cloud platform sends the control instruction to the energy storage system.

[0020] Furthermore, in step 2, the future photovoltaic power generation power P is calculated based on the historical operation data and the current operation data. fcstThe steps to making a prediction include:

[0021] S1-1. Data collection: Collect historical photovoltaic power P from the cloud platform hist , the current photovoltaic power P now ;

[0022] S1-2. Establishing the historical photovoltaic power database DP hist :Taking a single energy storage system as a unit, the local time T hist Converted into a vector, combined with the historical data of photovoltaic power generation corresponding to that time P hist , construct the historical photovoltaic power database DP using formula (1) hist :

[0023]

[0024] Among them, N represents the total amount of historical data of the current user, and each vector is sorted in chronological order;

[0025] S1-3. Establish the current vector: obtain the power generated by a user at the current moment With time information And establish the current vector

[0026] S1-4, prediction: Use the cosine similarity algorithm to convert the current vector With the historical photovoltaic power database DP hist The most similar vector is selected, and the photovoltaic power generation in one or more vectors following the most similar vector is used as the future photovoltaic power generation power P fcst The predicted value of

[0027] S1-5, data update, the current vector Update to the historical photovoltaic power database DP as the latest historical data hist middle.

[0028] Furthermore, in step 2, the future load power L is calculated based on the historical operation data and the current operation data. fcst The steps to making a prediction include:

[0029] S2-1. Data collection: Collect historical load L from the cloud platform hist and the current load L now , collect local holiday information based on the time of historical data H(T hist ), where H(T hist )=1 represents holidays, and the distinction between working days and rest days W(T hist ), other W(T hist )=1 represents a rest day;

[0030] S2-2. Establishing the historical load database DL hist : Combine historical load with workday and holiday information, model it at the 5-minute level, and generate a vector, recorded as TL hist =[T hist , H(T hist ), W(T hist )], and construct the historical load database DL using formula (2) hist :

[0031]

[0032] Where N represents the total number of historical data that the current user has;

[0033] S2-3. Establish the current vector: obtain the load power of a user at the current moment With time information And establish the current vector

[0034] S2-4, prediction: Use the cosine similarity algorithm to convert the current vector With the historical load power database DL hist The vectors in are compared, the most similar vector is selected, and the load power in one or more vectors following the most similar vector is used as the future load power L fcst The predicted value of

[0035] S2-5, data update, the current vector Update the historical load power database DL as the latest historical data hist middle.

[0036] Furthermore, the step of selecting a suitable operation mode using the output of the linear programming model in step 3 includes:

[0037] When P fcst -L fcst >=0, and P sell =0, and P charge >0, the strategy generation module selects the first mode; the first mode is that the energy storage system only uses photovoltaic power to charge the energy storage system;

[0038] When P buy >0, and P charge >0, the strategy generation module selects the second mode; the second mode is that the energy storage system charges the energy storage system in combination with the power of the power grid;

[0039] When P discharge , P chargeWhen both are 0, the strategy generation module selects the third mode, and the third mode is that the energy storage system remains in a static state;

[0040] When P discharge >0, and P sell =0, the strategy generation module selects the fourth mode, and the fourth mode is that the energy storage system only discharges to meet the demand of the load;

[0041] When P sell >0, and P discharge >0, the strategy generation module selects the fifth mode, and the fifth mode is that the energy storage system sells electricity back to the power grid to obtain revenue.

[0042] Furthermore, the constraints of the linear programming model include power balance constraints, charging and discharging power constraints, battery capacity constraints and power flow constraints;

[0043] The power balance constraint is:

[0044] P fcst +L fcst +PS load +P buy -P sell +P discharge +P charge +P waste =0

[0045] Among them, PS load Represents the load power during this period, P buy Represents the current power purchase, P sell Represents the electricity sold during this period, P discharge Represents the battery discharge power during this period, P charge Represents the battery charging power during this period, P waste Represents the amount of photovoltaic waste electricity;

[0046] The charging and discharging power constraints include: the battery charging and discharging power in each time period and the rated power satisfy:

[0047] PDMIN set ≤P discharge ≤PDMAX set ≤PD rated

[0048] PCMIN set ≤P charge ≤PCMAX set ≤PC rated

[0049] Where: PDMIN set Represents the set minimum discharge power, PDMAXset Represents the set maximum discharge power, PD reted Represents the rated discharge power of the battery, PCMIN set Represents the set minimum charging power, PCMAX set Represents the maximum charging power set, PC rated Represents the rated charging power of the battery, P discharge is the discharge power, P charge is the charging power;

[0050] The battery capacity constraint includes: the range of the remaining battery capacity in each time period, which satisfies the following formula:

[0051] 0≤SoCMIN set ≤SoC≤SoCMAX set ≤1

[0052] Among them SoCMIN set Represents the minimum battery SoC set, SoCMAX set represents the maximum battery SoC set. At the same time, in each period, P discharge and P charge Cannot cause SoC to be lower than SoCMIN set , exceeding SoCMAX set ;

[0053] The power flow constraint includes: the power flow between the power grid and the energy storage system remains unchanged in each time period, and the power flow of the battery remains unchanged. The formula is as follows:

[0054] GS*P sell +(1-GS)*P buy =0

[0055] BS*P discharge +(1-BS)*P charge =0

[0056] Among them, GS represents the grid control state in this period, and its value is 0 or 1; BS represents the battery control state in this period, and its value is 0 or 1; P sell For selling electricity.

[0057] Compared with the prior art, the present invention has the following advantages:

[0058] (1) This application models and predicts photovoltaic power, combines time information and local holiday information to predict the load of the energy storage system, and dynamically adjusts the control strategy of the energy storage system, thereby achieving refined power management and regulation, thereby effectively reducing users' electricity costs.

[0059] (2) This application abstracts five operation modes. Compared with directly issuing clear operation instructions, the algorithm of this application has higher robustness.

[0060] (3) The present application implements grid-level photovoltaic power generation prediction and household load prediction based on different users and regions, so that the prediction information has user characteristics, which is conducive to point-to-point regulation and control, and avoids the cost of training models for users in different regions separately.

[0061] (4) This application can effectively adapt to energy planning under the background of dynamic electricity prices, and maximize the user's economic benefits and save electricity bills by optimizing peak shaving and valley filling operations.

[0062] (5) This application only uses historical data when predicting photovoltaic power generation and load power, avoiding the use of weather factors and reducing reliance on multivariate data. BRIEF DESCRIPTION OF THE DRAWINGS

[0063] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0064] Figure 1 A schematic diagram of a generation system of a dynamic electricity price-based energy storage system control strategy according to this embodiment;

[0065] Figure 2 Schematic diagram of the structure of the prediction module of this embodiment. DETAILED DESCRIPTION

[0066] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention. Figure 1 FIG. 1 is a schematic diagram of a generation system 10 of a control strategy for an energy storage system 100 based on dynamic electricity prices in this embodiment. Figure 1 As shown, the generation system 10 includes a prediction module 101, a strategy generation module 102 and a cloud platform 103; the prediction module 101, the strategy generation module 102 and the cloud platform 103 are connected in sequence; and the cloud platform 13 is also connected to the prediction module 101.

[0067] Each module is now described in detail.

[0068] Prediction module 101, used to predict future photovoltaic power generation power P fcst and future load power L fcst .

[0069] Figure 2 FIG. 1 shows a schematic diagram of the structure of the prediction module 101 of this embodiment. Figure 2 The prediction module 101 shown includes a historical data acquisition module 1011, a first prediction module 1012, and a second prediction module 1013. Figure 1 , the historical data acquisition module 1011 is connected to the cloud platform 103; the first prediction module 1012 is connected to the historical data acquisition module 1011 and the strategy generation module 102 respectively; the second prediction module 1013 is connected to the historical data acquisition module 1011 and the strategy generation module 102 respectively.

[0070] The historical data acquisition module 1011 is used to acquire historical operation data and current operation data from the cloud platform 103; the historical operation data includes the historical photovoltaic power generation power P hist and historical load power L hist ; The current operating data includes the current photovoltaic power generation power P now and the current load power L fcst ;

[0071] The first prediction module 1012 is used to predict the future photovoltaic power generation power P in a certain time period in the future based on the historical operation data and the current operation data. fcst ;

[0072] The second prediction module 1013 is used to predict the future load power L in a certain period of time in the future based on the historical operation data and the current operation data. fcst ;

[0073] The first prediction module 1012 predicts the future photovoltaic power generation power P according to the historical operation data and the current operation data fcst The specific steps include:

[0074] S1-1. Data collection: Collect historical photovoltaic power P from the cloud platform hist , the current photovoltaic power P now ;

[0075] S1-2. Establishing the historical photovoltaic power database DP hist :Take a single energy storage system as a unit and convert the local time T hist Converted into a vector, combined with the historical data of photovoltaic power generation corresponding to that time P hist , construct the historical photovoltaic power database DP using formula (1) hist :

[0076]

[0077] Wherein, N=1000 represents that the energy storage system 100 has 1000 historical data vectors, and each vector is sorted in chronological order;

[0078] S1-3. Establish the current vector: obtain the power generated by a user at the current moment With time information And establish the current vector

[0079] S1-4, prediction: Use the cosine similarity algorithm to convert the current vector With the historical photovoltaic power database DP hist The most similar vector is selected, and the photovoltaic power generation in one or more vectors following the most similar vector is used as the future photovoltaic power generation power P fcst The predicted value of

[0080] S1-5, data update, the current vector Update to the historical photovoltaic power database DP as the latest historical data hist middle.

[0081] And, the second prediction module 1013 predicts the future load power L according to the historical operation data and the current operation data. fcst The steps include:

[0082] S2-1. Data collection: Collect historical load L from the cloud platform hist and the current load L now , collect local holiday information based on the time of historical data H(T hist ), where H(T hist )=1 represents holidays, and the distinction between working days and rest days W(T hist ), other W(T hist )=1 represents a rest day;

[0083] S2-2. Establishing the historical load database DL hist : Combine historical load with workday and holiday information, model it at the 5-minute level, and generate a vector, recorded as TL hist =[T hist , H(T hist ), W(T hist )], and construct the historical load database DL using formula (2) hist :

[0084]

[0085] Where N represents the total number of historical data that the current user has;

[0086] S2-3. Establish the current vector: obtain the load power of a user at the current moment With time information And establish the current vector

[0087] S2-4, prediction: Use the cosine similarity algorithm to convert the current vector With the historical load power database DL hist The vectors in the table are compared, the most similar vector is selected, and the load power in the next vector corresponding to the most similar vector is used as the future load power L fcst The predicted value of

[0088] S2-5, data update, the current vector Update the historical load power database DL as the latest historical data hist middle.

[0089] Strategy generation module 102, based on the future photovoltaic power generation power P obtained by prediction module 101 fcst and future load power L fcst , using the output of the linear programming model to select the appropriate operation mode and generate control instructions. The output includes P buy , P sell and P discharge , P charge ;

[0090] The linear programming model is:

[0091]

[0092] Where: T is the number of time periods for prediction, with H minutes as one time period, H is an integer, M sell is the electricity price for each time period, P sell is the electricity sold in each time period, M buy P is the electricity purchase price for each time period. buy The power purchased for each time period;

[0093] The steps for selecting the appropriate operating mode using the output of the linear programming model include:

[0094] When P fcst -L fcst >=0, and P sell =0, and P charge >0, the strategy generation module selects the first mode; the first mode is that the energy storage system 100 only uses photovoltaic power to charge the energy storage system 100;

[0095] When Pbuy >0, and P charge >0, the strategy generation module selects the second mode; the second mode is that the energy storage system 100 charges the energy storage system 100 in combination with the power of the power grid;

[0096] When P discharge , P charge When both are 0, the strategy generation module selects the third mode, and the third mode is that the energy storage system 100 remains in a static state;

[0097] When P discharge >0 and P sell =0, the strategy generation module selects the fourth mode, in which the energy storage system 100 only discharges to meet the load demand;

[0098] When P sell >0 and P discharge >0, the strategy generation module selects the fifth mode, and the fifth mode is that the energy storage system 100 sells electricity back to the power grid to obtain revenue.

[0099] The constraints of the linear programming model include power balance constraints, charging and discharging power constraints, battery capacity constraints, and power flow constraints;

[0100] The power balance constraint is:

[0101] P fcst +L fcst +PS load +P buy -P sell +P discharge +P charge +P waste =0

[0102] Among them, PS load Represents the load power during this period, P buy Represents the current power purchase, P sell Represents the electricity sold during this period, P discharge Represents the battery discharge power during this period, P charge Represents the battery charging power during this period, P waste Represents the amount of photovoltaic waste electricity;

[0103] The charging and discharging power constraints include: the battery charging and discharging power in each time period and the rated power meet the following requirements:

[0104] PDMIN set ≤P discharge ≤PDMAX set ≤PD rated

[0105] PCMIN set ≤P charge ≤PCMAX set ≤PC rated

[0106] Where: PDMIN set Represents the set minimum discharge power, PDMAX set Represents the set maximum discharge power, PD reted Represents the rated discharge power of the battery, PCMIN set Represents the set minimum charging power, PCMAX set Represents the maximum charging power set, PC rated Represents the rated charging power of the battery, P discharge is the discharge power, P charge is the charging power.

[0107] The battery capacity constraints include: the range of the remaining battery capacity in each time period, which satisfies the following formula:

[0108] 0≤SoCMIN set ≤SoC≤SoCMAX set ≤1

[0109] Among them SoCMIN set Represents the minimum battery SoC set, SoCMAX set represents the maximum battery SoC set. At the same time, in each period, P discharge and P charge Cannot cause SoC to be lower than SoCMIN set , exceeding SoCMAX set .

[0110] The power flow constraint includes: the power flow between the power grid and the energy storage system 100 remains unchanged in each time period, and the power flow of the battery remains unchanged. The formula is as follows:

[0111] GS*P sell +(1-GS)*P buy =0

[0112] BS*P discharge +(1-BS)*P charge =0

[0113] Among them, GS represents the grid control state in this period, and the value range is 0 or 1; BS represents the battery control state in this period, and the value range is 0 or 1; P sell For the power sold;

[0114] Optionally, this embodiment also includes localized constraints, including: the control strategy meets the management requirements of the location of the energy storage system 100. The management requirements can be local laws and regulations, or user rules and regulations. This application can adapt to policies and regulations in different regional contexts, meet the electricity needs of users in different countries and regions, and has better policy adaptability.

[0115] The cloud platform 103 is used to obtain the current operation data of the energy storage system 100 through the collector 104, store and update the historical data, and issue the control instructions to each energy storage system 100; Figure 1 As shown, the cloud platform 103 also collects operating data of the energy storage system 100 through the collector 104 .

[0116] On the other hand, based on the above device, this embodiment also provides a control strategy generation method for an energy storage system 100 based on dynamic electricity prices, comprising the following steps:

[0117] Step 1: Obtain historical operation data and current operation data of the energy storage system 100 from the cloud platform;

[0118] Step 2: Calculate the future photovoltaic power generation power P based on historical operation data and current operation data fcst Forecast and predict the future load power L based on historical operating data and current operating data fcst Make predictions;

[0119] Step 3: Based on the future photovoltaic power generation power P fcst 、Future load power L fcst and the output of the linear programming model, selecting a suitable operation mode for the energy storage system 100 and generating control instructions;

[0120] Step 4: The cloud platform sends the control instructions to the energy storage system 100.

[0121] The specific steps of step 2 and step 3 have been described in detail above and will not be repeated here.

[0122] The "equal" or "same" or "equal" disclosed in the present invention must take into account the distribution of engineering parameters, and the error distribution is within ±30%; the definition of "parallel" of two line segments or two straight lines is that the angle between the two line segments or two straight lines is less than or equal to 45 degrees; the definition of "perpendicular" of two line segments or two straight lines is that the angle between the two line segments or two straight lines is within the range of [60, 120] degrees; the definition of "phase mismatch" also needs to take into account the distribution of engineering parameters, and the error distribution of the degree of mismatch is within ±30%. In addition, in this article, relational terms such as first and second, etc. are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Moreover, the terms "include", "comprise" or any other variants thereof are intended to cover non-exclusive inclusion, so that a process, method, article or device including a series of elements includes not only those elements, but also includes other elements not explicitly listed, or also includes elements inherent to such process, method, article or device. Without more constraints, an element defined by the phrase "comprising a ..." does not exclude the existence of other identical elements in the process, method, article or apparatus comprising the element.

[0123] The various embodiments in this specification are described in a progressive manner, and each embodiment focuses on the differences from other embodiments. The same or similar parts between the various embodiments can be referenced to each other.

[0124] The above description of the disclosed embodiments enables one skilled in the art to implement or use the present invention. Various modifications to these embodiments will be apparent to one skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the present invention. Therefore, the present invention will not be limited to the embodiments shown herein, but rather to the widest scope consistent with the principles and novel features disclosed herein.

Claims

1. A control strategy generation system for an energy storage system based on dynamic electricity prices, characterized in that: include: Prediction module, strategy generation module and cloud platform; The prediction module includes a historical data acquisition module, a first prediction module and a second prediction module; the historical data acquisition module is connected to the first prediction module and the second prediction module respectively; the first prediction module is also connected to the strategy generation module; the second prediction module is also connected to the strategy generation module; the strategy generation module is also connected to the cloud platform; the cloud platform is also connected to the historical data acquisition module.

2. The control strategy generation system according to claim 1, characterized in that: The historical data acquisition module is used to obtain historical operation data and current operation data from the cloud platform; the historical operation data includes historical photovoltaic power generation power P hist and historical load power L hist The current operating data includes the current photovoltaic power generation power P now and the current load power L now ; The first prediction module is used to predict the future photovoltaic power generation power P in a certain time period in the future based on the historical operation data and the current operation data. fcst ; The second prediction module is used to predict the future load power L in a certain time period in the future based on the historical operation data and the current operation data. fcst ; The strategy generation module is used to generate a strategy based on the future photovoltaic power generation power P fcst and future load power L fcst , using the output of the linear programming model to select the appropriate operation mode and generate control instructions; The cloud platform is used to obtain the current operating data of the energy storage system through the collector, store and update the historical data, and issue the control instructions to each energy storage system.

3. The control strategy generation system according to claim 2, characterized in that: The output of the strategy generation module includes P buy , P sell , P discharge and P charge , the linear programming model is: Where: T is the number of time periods for prediction, with H minutes as one time period, H is an integer, M sell is the electricity price for each time period, P sell is the electricity sold in each time period, M buy P is the electricity purchase price for each time period. buy The power purchased in each time period, Maximize PROFIT is the maximum profit.

4. A control strategy generation method for an energy storage system based on dynamic electricity prices, using the control strategy generation system according to any one of claims 1 to 3, characterized in that: The following steps are involved: Step 1: Acquire the historical operation data and the current operation data of the energy storage system from the cloud platform; Step 2: Calculate the future photovoltaic power generation power P based on the historical operation data and the current operation data fcst According to the historical operation data and the current operation data, the future load power L is predicted. fcst Make predictions; Step 3: Based on the future photovoltaic power generation power P fcst 、Future load power L fcst and the output of the linear programming model, selecting a suitable operation mode for the energy storage system and generating a control instruction; Step 4: The cloud platform sends the control instruction to the energy storage system.

5. The generation method according to claim 4, characterized in that: In step 2, the future photovoltaic power generation power P is calculated based on the historical operation data and the current operation data. fcst The steps to making a prediction include: S1-1. Data collection: Collect historical photovoltaic power P from the cloud platform hist , the current photovoltaic power P now ; S1-2. Establishing the historical photovoltaic power database DP hist :Taking a single energy storage system as a unit, the local time T hist Converted into a vector, combined with the historical data of photovoltaic power generation corresponding to that time P hist , construct the historical photovoltaic power database DP using formula (1) hist : Among them, N represents the total amount of historical data of the current user, and each vector is sorted in chronological order; S1-3. Establish the current vector: obtain the power generated by a user at the current moment With time information And establish the current vector S1-4, prediction: Use the cosine similarity algorithm to convert the current vector With the historical photovoltaic power database DP hist The most similar vector is selected, and the photovoltaic power generation in one or more vectors following the most similar vector is used as the future photovoltaic power generation power P fcst The predicted value of S1-5, data update, the current vector Update to the historical photovoltaic power database DP as the latest historical data hist middle.

6. The generation method according to claim 4, characterized in that: In step 2, the future load power L is calculated based on the historical operation data and the current operation data. fcst The steps to making a prediction include: S2-1. Data collection: Collect historical load L from the cloud platform hist and the current load L now , collect local holiday information based on the time of historical data H(T hist ), where H(T hist )=1 represents holidays, and the distinction between working days and rest days W(T hist ), other W(T hist )=1 represents a rest day; S2-2. Establishing the historical load database DL hist : Combine historical load with workday and holiday information, model it at the 5-minute level, and generate a vector, recorded as TL hist =[T hist , H(T hist ), W(T hist )], and construct the historical load database DL using formula (2) hist : Where N represents the total number of historical data that the current user has; S2-3. Establish the current vector: obtain the load power of a user at the current moment With time information And establish the current vector S2-4, prediction: Use the cosine similarity algorithm to convert the current vector With the historical load power database DL hist The vectors in are compared, the most similar vector is selected, and the load power in one or more vectors following the most similar vector is used as the future load power L fcst The predicted value of S2-5, data update, the current vector Update the historical load power database DL as the latest historical data hist middle.

7. The generation method according to claim 4, characterized in that: The step of selecting a suitable operation mode using the output of the linear programming model in step 3 comprises: When P fcst -L fcst >=0, and P sell =0, and P charge >0, the strategy generation module selects the first mode; the first mode is that the energy storage system only uses photovoltaic power to charge the energy storage system; When P buy >0, and P charge >0, the strategy generation module selects the second mode; the second mode is that the energy storage system charges the energy storage system in combination with the power of the power grid; When P discharge , P charge When both are 0, the strategy generation module selects the third mode, and the third mode is that the energy storage system remains in a static state; When P discharge >0, and P sell =0, the strategy generation module selects the fourth mode, and the fourth mode is that the energy storage system only discharges to meet the demand of the load; When P sell >0, and P discharge >0, the strategy generation module selects the fifth mode, and the fifth mode is that the energy storage system sells electricity back to the power grid to obtain revenue.

8. The generation method according to claim 4, characterized in that: The constraints of the linear programming model include power balance constraints, charging and discharging power constraints, battery capacity constraints and power flow constraints; The power balance constraint is: P fcst +L fcst +PS load +P buy -P sell +P discharge +P charge +P waste =0 Among them, PS load Represents the load power during this period, P buy Represents the current power purchase, P sell Represents the electricity sales power during this period, P discharge Represents the battery discharge power during this period, P charge Represents the battery charging power during this period, P waste Represents the amount of photovoltaic waste electricity; The charging and discharging power constraints include: the battery charging and discharging power in each time period and the rated power satisfy: PDMIN set ≤P discharge ≤PDMAX set ≤PD rated PCMIN set ≤P charge ≤PCMAX set ≤PC rated Where: PDMIN set Represents the set minimum discharge power, PDMAX set Represents the set maximum discharge power, PD reted Represents the rated discharge power of the battery, PCMIN set Represents the set minimum charging power, PCMAX set Represents the maximum charging power set, PC rated Represents the rated charging power of the battery, P discharge is the discharge power, P charge is the charging power; The battery capacity constraint includes: the range of the remaining battery capacity in each time period, which satisfies the following formula: 0≤SoCMIN set ≤SoC≤SoCMAX set ≤1 Among them SoCMIN set Represents the minimum battery SoC set, SoCMAX set Represents the maximum battery SoC set. At the same time, in each period, P discharge and P charge Cannot cause SoC to be lower than SoCMIN set , exceeding SoCMAX set ; The power flow constraint includes: the power flow between the power grid and the energy storage system remains unchanged in each time period, and the power flow of the battery remains unchanged. The formula is as follows: GS*P sell +(1-GS)*P buy =0 BS*P discharge +(1-BS)*P charge =0 Among them, GS represents the grid control state in this period, and its value is 0 or 1; BS represents the battery control state in this period, and its value is 0 or 1; P sell For selling electricity.

Images