Photovoltaic building cluster hybrid integer linear optimization dispatching method and device based on tradable energy

Abstract

This invention relates to a mixed-integer linear optimization scheduling method for photovoltaic building clusters based on tradable energy, comprising: acquiring relevant equipment data of photovoltaic buildings; constructing scheduling response models for electric vehicles and flexible loads; constructing an optimal scheduling model for the photovoltaic building cluster based on the scheduling response models of electric vehicles and flexible loads; linearizing the nonlinear equations in the optimal scheduling model of the photovoltaic building cluster using the Big M method; substituting relevant equipment data and user data into the linearized optimal scheduling model of the photovoltaic building cluster, and obtaining the equipment output data when the objective function is optimal through optimization. This invention also discloses a mixed-integer linear optimization scheduling device for photovoltaic building clusters based on tradable energy. This invention applies the tradable energy system to the energy trading market of photovoltaic building clusters, simplifying the trading mode and reducing the demand for user data; it also allows the optimization model to be solved without complex iterative algorithms.

Description

Technical Field

[0001] This invention relates to the field of energy dispatching technology, and in particular to a hybrid integer linear optimization dispatching method and apparatus for photovoltaic building complexes based on tradable energy. Background Technology

[0002] Currently, electricity consumption in my country's building sector continues to grow. Photovoltaic buildings, represented by "photovoltaic-storage-DC-flexible", are an important development direction for electrification in the building sector. They have changed the positioning of buildings in the power system and transformed the traditional building power system into a new type of building power system that integrates energy production, energy consumption, and energy storage.

[0003] With the large-scale integration of adjustable power sources such as photovoltaic (PV) power, energy storage devices, electric vehicles, and flexible loads into PV buildings, the fluctuations in the output of these devices will have a significant impact on the power grid's supply and demand balance. Furthermore, traditional electricity market trading platforms are not suitable for the needs and development of PV buildings. Therefore, ensuring that adjustable power sources of different types and sizes can be integrated flexibly and amicably, and that they have equal trading rights in the electricity market, is a pressing issue that needs to be addressed in the current electricity trading market.

[0004] Most current optimization scheduling schemes in the power trading market are nonlinear optimization problems, requiring complex iterative algorithms for solution. Common iterative algorithms include genetic algorithms, particle swarm optimization, and wolf pack algorithms. However, iterative algorithms suffer from drawbacks such as low accuracy of optimization results, long iteration times, and slow solution speed. Therefore, it is necessary to linearize nonlinear optimization problems to make them solvable by commercial solvers. Current research on linearization of nonlinear optimization problems mostly focuses on introducing slack variables and selecting new basic variables, resulting in complex solution processes. Summary of the Invention

[0005] To address the challenges of traditional market trading platforms failing to accommodate large-scale integration of small-volume adjustable resources such as photovoltaics, energy storage, electric vehicles, and flexible loads, as well as the computational complexity and solution difficulties arising from high dimensionality and nonlinearity in existing scheduling methods, the primary objective of this invention is to provide a hybrid integer linear optimization scheduling method for photovoltaic building clusters based on tradable energy. This method utilizes tradable energy systems as the framework for a photovoltaic building energy trading market, uses price as a parameter to adjust the global supply and demand dynamic balance of the system, optimizes the output of each device, and simplifies the trading model.

[0006] To achieve the above objectives, the present invention adopts the following technical solution: a mixed-integer linear optimization scheduling method for photovoltaic building clusters based on tradable energy, the method comprising the following sequential steps:

[0007] (1) Obtain relevant equipment data for photovoltaic buildings, including hourly output data of building photovoltaics, hourly predicted power of rigid loads and flexible loads;

[0008] (2) Construct scheduling response models for electric vehicles and flexible loads;

[0009] (3) Based on the scheduling response model of electric vehicles and flexible loads, construct an optimized scheduling model for photovoltaic building clusters;

[0010] (4) Linearize the nonlinear equations in the photovoltaic building cluster optimization scheduling model based on the Big M method to obtain the photovoltaic building cluster optimization scheduling model after linearization.

[0011] (5) Substitute the relevant equipment data and user data into the photovoltaic building cluster optimization scheduling model after linearization, and obtain the equipment output data when the objective function is optimal through optimization solution.

[0012] Step (2) specifically refers to: the charging power of the electric vehicle is:

[0013] (1)

[0014] in, This represents the power reduction of electric vehicle k in building b during scheduling at time t; This represents the rated charging power of electric vehicle k in building b; This represents the actual charging power of electric vehicle k in building b at time t;

[0015] Construct a functional relationship between the power reduction of electric vehicles and the compensation price:

[0016] (2)

[0017] in: This represents the actual compensation price of electric vehicle k in building b at time t; This represents the compensation price when the charging power of electric vehicle k in building b is 0. This represents the compensation price for electric vehicle k in building b when it discharges at its rated power.

[0018] Calculate the owner's income when the electric vehicle is discharging at its rated power. :

[0019] (3)

[0020] In the formula: Indicates the charging efficiency of electric vehicles; This represents the wear compensation coefficient of an electric vehicle battery. Indicates the electricity price at time t; Indicates the duration of each time step;

[0021] Calculate the owner's expenses when an electric vehicle discharges at its rated power. :

[0022] (4)

[0023] The compensation price for electric vehicles discharging at rated power is calculated using equations (3) and (4). :

[0024] (5)

[0025] The power of flexible loads participating in power grid dispatch is:

[0026] (6)

[0027] in, Let x represent the power of the flexible load x that the i-th user of building b responds to at time t. When x=1, it is a transferable load; when x=2, it is a shiftable load; and when x=3, it is a load that can be reduced. This represents the power of the flexible load x of the i-th user in building b at time t before scheduling. Let x represent the power of the flexible load x of the i-th user in building b at time t after scheduling;

[0028] The power of the flexible load response is a function of the compensation price, therefore this function is expressed as:

[0029] (7)

[0030] in, This represents the compensation price for the actual flexible load x of the i-th user in building b at time t; Indicates when At that time, the compensation price for the flexible load x expected by the i-th user of building b;

[0031] Equations (2) and (7) constitute the scheduling response model for electric vehicles and flexible loads.

[0032] Step (3) specifically refers to the objective function of the photovoltaic building cluster optimization scheduling model being:

[0033] (8)

[0034] in, This indicates the cost of operating and maintaining the equipment; This indicates the cost of flexible load response compensation. This indicates the compensation cost for the electric vehicle response; This indicates the cost of electricity purchased by the building complex from the power grid;

[0035] Equipment operation and maintenance costs As shown in equation (9):

[0036] (9)

[0037] Where B represents the number of buildings in the building complex; T represents the total system operating time; This indicates the operating and maintenance cost per unit capacity of photovoltaic panels; This represents the output of the photovoltaic panels on building b at time t; This indicates the operating and maintenance cost per unit capacity of the energy storage device; This represents the charging power of the energy storage device in building b at time t; This represents the discharge power of the energy storage device in building b at time t;

[0038] The power of flexible loads participating in power grid dispatch is calculated from the absolute value of the difference between the power of flexible loads before and after dispatch. The power change is calculated twice; therefore, the compensation cost for flexible load dispatch is calculated. for:

[0039] (10)

[0040] Where I represents the total number of users participating in flexible load dispatch in building b; This represents the compensation price for the actual flexible load x of the i-th user in building b at time t; Let x represent the power of the flexible load x that the i-th user of building b responds to at time t;

[0041] Electric vehicle dispatch compensation fee for:

[0042] (11)

[0043] Where K represents the total number of electric vehicles in building b; This represents the actual compensation price of electric vehicle k in building b at time t; This represents the power reduction of electric vehicle k in building b during scheduling at time t;

[0044] Electricity purchase cost for:

[0045] (12)

[0046] in, This represents the amount of electricity that building b purchases from the power grid at time t.

[0047] The constraints of the photovoltaic building cluster optimization scheduling model include photovoltaic building cluster power balance constraints, energy storage device constraints, electric vehicle constraints, and flexible load constraints.

[0048] The power balance constraint of the photovoltaic building complex is shown in equation (13):

[0049] (13)

[0050] in, This represents the power of interaction between building b and other buildings at time t; This represents the rigid load power of building b at time t; Let x represent the power of the flexible load x of the i-th user in building b at time t after scheduling; This represents the actual charging power of electric vehicle k in building b at time t;

[0051] The constraints of the energy storage device are shown in equation (14):

[0052] (14)

[0053] In the formula, and These are the upper and lower limits of the battery charging power, respectively. and These are the upper and lower limits of the battery's discharge power, respectively.

[0054] The electric vehicle constraints are shown in equation (15):

[0055] (15)

[0056] In the formula, Let SOC be the battery SOC level of electric vehicle k at time t; and These are the upper and lower limits of the State of Charge (SOC) for electric vehicle batteries, respectively. The time it takes for electric car k to leave the building; The time it takes for electric car k to arrive at the building;

[0057] The flexible load constraint is shown in equation (16):

[0058] (16).

[0059] Step (4) specifically includes the following steps:

[0060] (4a) The function relationship between the power reduction of electric vehicles and the compensation price, i.e., equation (2), is linearized using the Big M method:

[0061] (2)

[0062] In the formula, This represents the actual compensation price of electric vehicle k in building b at time t; This represents the compensation price when the charging power of electric vehicle k in building b is 0. This represents the compensation price for electric vehicle k in building b when it discharges at its rated power.

[0063] Introduce 3 binary variables , and Equation (2) can be transformed into:

[0064] (17)

[0065] (18)

[0066] (19)

[0067] (20)

[0068] Where M is a sufficiently large positive number; This represents the power reduction of electric vehicle k in building b during scheduling at time t; This represents the rated charging power of electric vehicle k in building b; This represents the rated charging power of electric vehicle k in building b; This represents the compensation price when the charging power of electric vehicle k in building b is 0. This represents the compensation price for electric vehicle k in building b when it discharges at its rated power.

[0069] Equation (17) contains a nonlinear term consisting of the product of a binary variable and a continuous variable. To linearize it, two auxiliary continuous variables are introduced. and :

[0070] (twenty one)

[0071] (twenty two)

[0072] Equation (17) is then transformed into equation (23):

[0073] (twenty three)

[0074] Auxiliary continuous variables and Constraints must be met:

[0075] (twenty four)

[0076] (25)

[0077] (4b) Using the Big M method to compensate for flexible load response costs That is, linearize equation (10):

[0078] (10)

[0079] Where I represents the total number of users participating in flexible load dispatch in building b; This represents the compensation price for the actual flexible load x of the i-th user in building b at time t; Let x represent the power of the flexible load x that the i-th user of building b responds to at time t;

[0080] and The product of these terms forms a nonlinear term. To linearize it, a binary variable is introduced. Define a set Find the optimal solution, and locate the optimal solution:

[0081] (26)

[0082] (27)

[0083] Where J represents the total number of elements that can be selected from the set; The maximum allowable compensation price is x for flexible load;

[0084] Define a new variable To satisfy Then equation (10) is transformed into:

[0085] (28)

[0086] The following constraints must be met:

[0087] (29)

[0088] (4c) Using the Big M method to calculate electric vehicle dispatch compensation costs That is, linearize equation (11):

[0089] (11)

[0090] Where K represents the total number of electric vehicles in building b; This represents the actual compensation price of electric vehicle k in building b at time t; This represents the power reduction of electric vehicle k in building b during scheduling at time t;

[0091] and The product of these terms forms a nonlinear term, introducing a binary variable. Define a set Find the optimal solution and locate the optimal solution:

[0092] (30)

[0093] (31)

[0094] In the formula, The maximum allowable compensation price for electric vehicles;

[0095] Define a new variable To satisfy Then equation (10) is transformed into:

[0096] (32)

[0097] In the formula, T represents the total system running time;

[0098] The following constraints must be met:

[0099] (33)

[0100] Another objective of this invention is to provide a hybrid integer linear optimization scheduling device for photovoltaic building complexes based on tradable energy, comprising:

[0101] The data acquisition module is used to acquire relevant equipment data, including hourly output data of building photovoltaic systems and hourly predicted power of rigid and flexible loads.

[0102] The photovoltaic building cluster modeling module is used to construct the optimal scheduling model, objective function, and constraints for photovoltaic building clusters.

[0103] The calculation and solution module is used to linearize the photovoltaic building cluster optimization scheduling model, substitute various parameters into the pre-constructed photovoltaic building cluster optimization scheduling model, and obtain the photovoltaic building cluster optimization scheduling scheme with the minimum objective function through optimization solution.

[0104] As can be seen from the above technical solution, the beneficial effects of the present invention are as follows: First, the present invention applies the tradable energy system to the energy trading market of photovoltaic building clusters, improving the problem that existing energy trading methods cannot adapt to the large-scale access of small-volume adjustable resources; Second, the electric vehicle scheduling response strategy proposed in the present invention simplifies the trading mode and reduces the demand for user data by considering the individual differences of electric vehicle owners; Third, the present invention transforms the complex nonlinear problem into a mixed integer linear programming problem, so that the optimization model can be solved without the need for complex iterative algorithms. Attached Figure Description

[0105] Figure 1 This is a flowchart of the method of the present invention;

[0106] Figure 2 This is a basic structural diagram of a photovoltaic building;

[0107] Figure 3 This is a structural diagram of the energy trading market for photovoltaic building complexes based on tradable energy systems;

[0108] Figure 4 This is a graph showing the relationship between the reduced charging power of electric vehicles and the compensation price. Detailed Implementation

[0109] like Figure 1 As shown, a mixed-integer linear optimization scheduling method for photovoltaic building clusters based on tradable energy includes the following sequential steps:

[0110] (1) Obtain relevant equipment data for photovoltaic buildings, including hourly output data of building photovoltaics, hourly predicted power of rigid loads and flexible loads;

[0111] (2) Construct scheduling response models for electric vehicles and flexible loads;

[0112] (3) Based on the scheduling response model of electric vehicles and flexible loads, construct an optimized scheduling model for photovoltaic building clusters;

[0113] (4) Linearize the nonlinear equations in the photovoltaic building cluster optimization scheduling model based on the Big M method to obtain the photovoltaic building cluster optimization scheduling model after linearization.

[0114] (5) Substitute the relevant equipment data and user data into the photovoltaic building cluster optimization scheduling model after linearization, and obtain the equipment output data when the objective function is optimal through optimization solution.

[0115] Step (2) specifically refers to: the charging power of the electric vehicle is:

[0116] (1)

[0117] in, This represents the power reduction of electric vehicle k in building b during scheduling at time t; This represents the rated charging power of electric vehicle k in building b; This represents the actual charging power of electric vehicle k in building b at time t;

[0118] like Figure 4 As shown, a functional relationship is constructed between the power reduction of electric vehicles and the compensation price:

[0119] (2)

[0120] in: This represents the actual compensation price of electric vehicle k in building b at time t; This represents the compensation price when the charging power of electric vehicle k in building b is 0. This represents the compensation price for electric vehicle k in building b when it discharges at its rated power.

[0121] Calculate the owner's income when the electric vehicle is discharging at its rated power. :

[0122] (3)

[0123] In the formula: Indicates the charging efficiency of electric vehicles; This represents the wear compensation coefficient of an electric vehicle battery. Indicates the electricity price at time t; Indicates the duration of each time step;

[0124] Calculate the owner's expenses when an electric vehicle discharges at its rated power. :

[0125] (4)

[0126] The compensation price for electric vehicles discharging at rated power is calculated using equations (3) and (4). :

[0127] (5)

[0128] The power of flexible loads participating in power grid dispatch is:

[0129] (6)

[0130] in, Let x represent the power of the flexible load x that the i-th user of building b responds to at time t. When x=1, it is a transferable load; when x=2, it is a shiftable load; and when x=3, it is a load that can be reduced. This represents the power of the flexible load x of the i-th user in building b at time t before scheduling. Let x represent the power of the flexible load x of the i-th user in building b at time t after scheduling;

[0131] The power of the flexible load response is a function of the compensation price, therefore this function is expressed as:

[0132] (7)

[0133] in, This represents the compensation price for the actual flexible load x of the i-th user in building b at time t; Indicates when At that time, the compensation price for the flexible load x expected by the i-th user of building b;

[0134] Equations (2) and (7) constitute the scheduling response model for electric vehicles and flexible loads.

[0135] Step (3) specifically refers to the objective function of the photovoltaic building cluster optimization scheduling model being:

[0136] (8)

[0137] in, This indicates the cost of operating and maintaining the equipment; This indicates the cost of flexible load response compensation. This indicates the compensation cost for the electric vehicle response; This indicates the cost of electricity purchased by the building complex from the power grid;

[0138] Equipment operation and maintenance costs As shown in equation (9):

[0139] (9)

[0140] Where B represents the number of buildings in the building complex; T represents the total system operating time; This indicates the operating and maintenance cost per unit capacity of photovoltaic panels; This represents the output of the photovoltaic panels on building b at time t; This indicates the operating and maintenance cost per unit capacity of the energy storage device; This represents the charging power of the energy storage device in building b at time t; This represents the discharge power of the energy storage device in building b at time t;

[0141] The power of flexible loads participating in power grid dispatch is calculated from the absolute value of the difference between the power of flexible loads before and after dispatch. The power change is calculated twice; therefore, the compensation cost for flexible load dispatch is calculated. for:

[0142] (10)

[0143] Where I represents the total number of users participating in flexible load dispatch in building b; This represents the compensation price for the actual flexible load x of the i-th user in building b at time t; Let x represent the power of the flexible load x that the i-th user of building b responds to at time t;

[0144] Electric vehicle dispatch compensation fee for:

[0145] (11)

[0146] Where K represents the total number of electric vehicles in building b; This represents the actual compensation price of electric vehicle k in building b at time t; This represents the power reduction of electric vehicle k in building b during scheduling at time t;

[0147] Electricity purchase cost for:

[0148] (12)

[0149] in, This represents the amount of electricity that building b purchases from the power grid at time t.

[0150] The constraints of the photovoltaic building cluster optimization scheduling model include photovoltaic building cluster power balance constraints, energy storage device constraints, electric vehicle constraints, and flexible load constraints.

[0151] The power balance constraint of the photovoltaic building complex is shown in equation (13):

[0152] (13)

[0153] in, This represents the power of interaction between building b and other buildings at time t; This represents the rigid load power of building b at time t; Let x represent the power of the flexible load x of the i-th user in building b at time t after scheduling; This represents the actual charging power of electric vehicle k in building b at time t;

[0154] The constraints of the energy storage device are shown in equation (14):

[0155] (14)

[0156] In the formula, and These are the upper and lower limits of the battery charging power, respectively. and These are the upper and lower limits of the battery's discharge power, respectively.

[0157] The electric vehicle constraints are shown in equation (15):

[0158] (15)

[0159] In the formula, Let SOC be the battery SOC level of electric vehicle k at time t; and These are the upper and lower limits of the State of Charge (SOC) for electric vehicle batteries, respectively. The time it takes for electric car k to leave the building; The time it takes for electric car k to arrive at the building;

[0160] The flexible load constraint is shown in equation (16):

[0161] (16).

[0162] Step (4) specifically includes the following steps:

[0163] (4a) The function relationship between the power reduction of electric vehicles and the compensation price, i.e., equation (2), is linearized using the Big M method:

[0164] (2)

[0165] In the formula, This represents the actual compensation price of electric vehicle k in building b at time t; This represents the compensation price when the charging power of electric vehicle k in building b is 0. This represents the compensation price for electric vehicle k in building b when it discharges at its rated power.

[0166] Introduce 3 binary variables , and Equation (2) can be transformed into:

[0167] (17)

[0168] (18)

[0169] (19)

[0170] (20)

[0171] Where M is a sufficiently large positive number; This represents the power reduction of electric vehicle k in building b during scheduling at time t; This represents the rated charging power of electric vehicle k in building b; This represents the rated charging power of electric vehicle k in building b; This represents the compensation price when the charging power of electric vehicle k in building b is 0. This represents the compensation price for electric vehicle k in building b when it discharges at its rated power.

[0172] Equation (17) contains a nonlinear term consisting of the product of a binary variable and a continuous variable. To linearize it, two auxiliary continuous variables are introduced. and :

[0173] (twenty one)

[0174] (twenty two)

[0175] Equation (17) is then transformed into equation (23):

[0176] (twenty three)

[0177] Auxiliary continuous variables and Constraints must be met:

[0178] (twenty four)

[0179] (25)

[0180] (4b) Using the Big M method to compensate for flexible load response costs That is, linearize equation (10):

[0181] (10)

[0182] Where I represents the total number of users participating in flexible load dispatch in building b; This represents the compensation price for the actual flexible load x of the i-th user in building b at time t; Let x represent the power of the flexible load x that the i-th user of building b responds to at time t;

[0183] and The product of these terms forms a nonlinear term. To linearize it, a binary variable is introduced. Define a set Find the optimal solution, and locate the optimal solution:

[0184] (26)

[0185] (27)

[0186] Where J represents the total number of elements that can be selected from the set; The maximum allowable compensation price is x for flexible load;

[0187] Define a new variable To satisfy Then equation (10) is transformed into:

[0188] (28)

[0189] The following constraints must be met:

[0190] (29)

[0191] (4c) Using the Big M method to calculate electric vehicle dispatch compensation costs That is, linearize equation (11):

[0192] (11)

[0193] Where K represents the total number of electric vehicles in building b; This represents the actual compensation price of electric vehicle k in building b at time t; This represents the power reduction of electric vehicle k in building b during scheduling at time t;

[0194] and The product of these terms forms a nonlinear term, introducing a binary variable. Define a set Find the optimal solution and locate the optimal solution:

[0195] (30)

[0196] (31)

[0197] In the formula, The maximum allowable compensation price for electric vehicles;

[0198] Define a new variable To satisfy Then equation (10) is transformed into:

[0199] (32)

[0200] In the formula, T represents the total system running time;

[0201] The following constraints must be met:

[0202] (33)

[0203] This device includes:

[0204] The data acquisition module is used to acquire relevant equipment data, including hourly output data of building photovoltaic systems and hourly predicted power of rigid and flexible loads.

[0205] The photovoltaic building cluster modeling module is used to construct the optimal scheduling model, objective function, and constraints for photovoltaic building clusters.

[0206] The calculation and solution module is used to linearize the photovoltaic building cluster optimization scheduling model, substitute various parameters into the pre-constructed photovoltaic building cluster optimization scheduling model, and obtain the photovoltaic building cluster optimization scheduling scheme with the minimum objective function through optimization solution.

[0207] The linearized photovoltaic building cluster optimization scheduling model was compiled using the Yalmip toolbox and solved using the Gurobi or Cplex solver.

[0208] Taking a residential building in a community that includes photovoltaic power sources, energy storage devices, electric vehicles, and flexible loads as the research object, such as... Figure 2 As shown in Table 1, the parameters of each piece of equipment in the building are shown in Table 1, and the electricity price adopts time-of-use pricing.

[0209] Table 1 Main parameters of various equipment in the building

[0210]

[0211] A simulation analysis was conducted on a photovoltaic building complex consisting of three buildings, examining the information and energy flow relationships between the photovoltaic building complex, the power distribution network, the trading market, and the users. Figure 3 As shown, the energy storage capacity of each building is 500 kWh, and the photovoltaic capacities are 1000 kW, 600 kW, and 500 kW, respectively. The simulation scheduling cycle is set to 1 day. The optimization results are shown in Table 2. Table 2 shows that the operating and maintenance costs of the equipment, as well as the response compensation costs of flexible loads and electric vehicles, are much lower than the cost of purchasing electricity. However, through the scheduling of flexible loads and electric vehicles, the building's electricity load curve is optimized, better matching the output curve of the photovoltaic equipment and reducing the pressure on the power grid. In the inter-building electricity trading process, Building 1 sells electricity to Buildings 2 and 3, fully utilizing the output of the photovoltaic equipment and reducing the building's demand for energy storage equipment.

[0212] Table 2 Optimization Results

[0213]

[0214] In summary, this invention applies tradable energy systems to the energy trading market of photovoltaic building complexes, improving the existing energy trading methods' inability to adapt to the large-scale access of small-volume adjustable resources; the electric vehicle scheduling response strategy proposed in this invention simplifies the trading model and reduces the demand for user data by considering the individual differences of electric vehicle owners; this invention transforms complex nonlinear problems into mixed-integer linear programming problems, enabling the optimization model to be solved without the need for complex iterative algorithms.

Claims

1. A mixed-integer linear optimization scheduling method for photovoltaic building clusters based on tradable energy, characterized in that: The method includes the following steps in sequence: (1) Obtain relevant equipment data for photovoltaic buildings, including hourly output data of building photovoltaics, hourly predicted power of rigid loads and flexible loads; (2) Construct scheduling response models for electric vehicles and flexible loads; establish a functional relationship between the power reduction of electric vehicles and the compensation price: (2) in: This represents the actual compensation price of electric vehicle k in building b at time t; This represents the compensation price when the charging power of electric vehicle k in building b is 0. This represents the compensation price for electric vehicle k in building b when it discharges at its rated power. This represents the rated charging power of electric vehicle k in building b; The power of the flexible load response is a function of the compensation price, therefore this function is expressed as: (7) in, This represents the compensation price for the actual flexible load x of the i-th user in building b at time t; Indicates when At that time, the compensation price for the flexible load x expected by the i-th user of building b; Let x represent the power of the flexible load x of the i-th user in building b at time t before scheduling; Equations (2) and (7) constitute the dispatch response model for electric vehicles and flexible loads; (3) Based on the scheduling response model of electric vehicles and flexible loads, an optimal scheduling model for photovoltaic building clusters is constructed; the objective function of the optimal scheduling model for photovoltaic building clusters is: (8) in, This indicates the cost of operating and maintaining the equipment; This indicates the cost of flexible load response compensation. This indicates the compensation cost for the electric vehicle response; This indicates the cost of electricity purchased by the building complex from the power grid; The constraints of the photovoltaic building cluster optimization scheduling model include photovoltaic building cluster power balance constraints, energy storage device constraints, electric vehicle constraints, and flexible load constraints. (4) The nonlinear equations in the photovoltaic building cluster optimization scheduling model are linearized based on the Big M method to obtain the linearized photovoltaic building cluster optimization scheduling model; the nonlinear equations include the functional relationship between the power reduction of electric vehicles and the compensation price, and the flexible load response compensation cost. Electric vehicle dispatch compensation fees ; (5) Substitute the relevant equipment data and user data into the photovoltaic building cluster optimization scheduling model after linearization, and obtain the equipment output data when the objective function is optimal through optimization solution.

2. The hybrid integer linear optimization scheduling method for photovoltaic building clusters based on tradable energy as described in claim 1, characterized in that: Step (2) specifically refers to: the charging power of the electric vehicle is: (1) in, This represents the power reduction of electric vehicle k in building b during scheduling at time t; This represents the actual charging power of electric vehicle k in building b at time t; Calculate the owner's income when the electric vehicle is discharging at its rated power. : (3) In the formula: Indicates the charging efficiency of electric vehicles; This represents the wear compensation coefficient of an electric vehicle battery. Indicates the electricity price at time t; Indicates the duration of each time step; Calculate the owner's expenses when an electric vehicle discharges at its rated power. : (4) The compensation price for electric vehicles discharging at rated power is calculated using equations (3) and (4). : (5) The power of flexible loads participating in power grid dispatch is: (6) in, Let x represent the power of the flexible load x that the i-th user of building b responds to at time t. When x=1, it is a transferable load; when x=2, it is a shiftable load; and when x=3, it is a load that can be reduced. Let x represent the power of the flexible load x of the i-th user in building b at time t after scheduling.

3. The mixed-integer linear optimization scheduling method for photovoltaic building clusters based on tradable energy as described in claim 1, characterized in that: Step (3) specifically refers to: equipment operation and maintenance costs. As shown in equation (9): (9) Where B represents the number of buildings in the building complex; T represents the total system operating time; This indicates the operating and maintenance cost per unit capacity of photovoltaic panels; This represents the output of the photovoltaic panels on building b at time t; This indicates the operating and maintenance cost per unit capacity of the energy storage device; This represents the charging power of the energy storage device in building b at time t; This represents the discharge power of the energy storage device in building b at time t; The power of flexible loads participating in power grid dispatch is calculated from the absolute value of the difference between the power of flexible loads before and after dispatch. The power change is calculated twice; therefore, the compensation cost for flexible load dispatch is calculated. for: (10) Where I represents the total number of users participating in flexible load dispatch in building b; This represents the compensation price for the actual flexible load x of the i-th user in building b at time t; Let x represent the power of the flexible load x that the i-th user of building b responds to at time t; Electric vehicle dispatch compensation fee for: (11) Where K represents the total number of electric vehicles in building b; This represents the actual compensation price of electric vehicle k in building b at time t; This represents the power reduction of electric vehicle k in building b during scheduling at time t; Electricity purchase cost for: (12) in, This represents the amount of electricity that building b purchases from the power grid at time t. The power balance constraint of the photovoltaic building complex is shown in equation (13): (13) in, This represents the power of interaction between building b and other buildings at time t; This represents the rigid load power of building b at time t; Let x represent the power of the flexible load x of the i-th user in building b at time t after scheduling; This represents the actual charging power of electric vehicle k in building b at time t; The constraints of the energy storage device are shown in equation (14): (14) In the formula, and These are the upper and lower limits of the battery charging power, respectively. and These are the upper and lower limits of the battery's discharge power, respectively. The electric vehicle constraints are shown in equation (15): (15) In the formula, Let SOC be the battery SOC level of electric vehicle k at time t; and These are the upper and lower limits of the State of Charge (SOC) for electric vehicle batteries, respectively. The time it takes for electric car k to leave the building; The time it takes for electric car k to arrive at the building; The flexible load constraint is shown in equation (16): (16)。 4. The mixed-integer linear optimization scheduling method for photovoltaic building clusters based on tradable energy as described in claim 1, characterized in that: Step (4) specifically includes the following steps: (4a) The function relationship between the power reduction of electric vehicles and the compensation price, i.e., equation (2), is linearized using the Big M method: (2) Introduce 3 binary variables , and Equation (2) can be transformed into: (17) (18) (19) (20) Where M is a sufficiently large positive number; This represents the rated charging power of electric vehicle k in building b; This represents the rated charging power of electric vehicle k in building b; This represents the compensation price when the charging power of electric vehicle k in building b is 0. This represents the compensation price for electric vehicle k in building b when it discharges at its rated power. Equation (17) contains a nonlinear term consisting of the product of a binary variable and a continuous variable. To linearize it, two auxiliary continuous variables are introduced. and : (21) (22) Equation (17) is then transformed into equation (23): (23) Auxiliary continuous variables and Constraints must be met: (24) (25) (4b) Using the Big M method to compensate for flexible load response costs That is, linearize equation (10): (10) Where I represents the total number of users participating in flexible load dispatch in building b; This represents the compensation price for the actual flexible load x of the i-th user in building b at time t; Let x represent the power of the flexible load x that the i-th user of building b responds to at time t; and The product of these terms forms a nonlinear term. To linearize it, a binary variable is introduced. Define a set Find the optimal solution, and locate the optimal solution: (26) (27) Where J represents the total number of elements that can be selected from the set; The maximum allowable compensation price is x for flexible load; Define a new variable To satisfy Then equation (10) becomes: (28) The following constraints must be met: (29) (4c) Using the Big M method to calculate electric vehicle dispatch compensation costs That is, linearize equation (11): (11) Where K represents the total number of electric vehicles in building b; This represents the actual compensation price of electric vehicle k in building b at time t; This represents the power reduction of electric vehicle k in building b during scheduling at time t; and The product of these terms forms a nonlinear term, introducing a binary variable. Define a set Find the optimal solution and locate the optimal solution: (30) (31) In the formula, The maximum allowable compensation price for electric vehicles; Define a new variable To satisfy Then equation (10) becomes: (32) In the formula, T represents the total system running time; The following constraints must be met: (33)。 5. An apparatus for implementing the hybrid integer linear optimization scheduling method for photovoltaic building clusters based on tradable energy as described in any one of claims 1 to 4, characterized in that: include: The data acquisition module is used to acquire relevant equipment data, including hourly output data of building photovoltaic systems and hourly predicted power of rigid and flexible loads. The photovoltaic building cluster modeling module is used to construct the optimal scheduling model, objective function, and constraints for photovoltaic building clusters. The calculation and solution module is used to linearize the photovoltaic building cluster optimization scheduling model, substitute various parameters into the pre-constructed photovoltaic building cluster optimization scheduling model, and obtain the photovoltaic building cluster optimization scheduling scheme with the minimum objective function through optimization solution.

Images