An optimization and coordination method for multi-layer heterogeneous smart energy systems

Abstract

The present invention discloses an optimization and coordination method for a multi-layer heterogeneous smart energy system, which belongs to the field of power systems. The present invention can reduce the operating cost of the smart energy system, improve the operating efficiency and reliability of the smart energy system, and attract different user groups to provide related services, thereby improving the economy and safety of the power system. The method includes constructing an upper optimization model, including load rewards, user risk tolerance, electricity purchase and sales prices, power, network, and distributed power output models and constraints of the smart energy system, and constructing a lower optimization model, including power balance, distributed power output, flexible load, and energy storage related models and constraints; after the upper model and the lower model are established respectively, the upper and lower optimization models are reconstructed to obtain analytical solutions, so that the models can be directly calculated. The present invention can coordinate the efficient, economical, and green operation of various resources, thereby achieving the purpose of efficient resource calculation and optimization scheduling.

Description

Technical Field

[0001] The present invention belongs to the field of power systems, and in particular relates to an optimization and coordination method for a multi-layer heterogeneous smart energy system. Background Art

[0002] Smart energy system is a modern energy system that integrates and optimizes energy production, transmission, distribution and consumption. It uses advanced information and communication technology, Internet of Things, smart grid technology, big data analysis and artificial intelligence to achieve efficient, reliable, clean and sustainable development of energy system. The core goal of smart energy system is to improve energy efficiency, reduce costs, enhance system flexibility and resilience through intelligent management and control, and promote the integration and utilization of renewable energy. The optimization and coordination of smart energy system involves all aspects of energy production, transmission, distribution and consumption, and its goal is to achieve efficient utilization of energy resources and economic operation of the system. In smart energy system, a large number of electric vehicles, energy storage, load and other resources are included. The optimization and coordination of various resources is not only crucial to improve the operation efficiency and reliability of the system; however, the direct participation of multiple user groups in the clearing of power system will increase the burden of system optimization calculation or cause no solution, and may endanger the safety of user groups decomposing the results of optimization calculation instructions to individual users, resulting in reduced system security. Summary of the invention

[0003] Based on the above shortcomings, the present invention proposes an optimization and coordinated operation method for a multi-layer heterogeneous smart energy system, which can reduce the operating cost of the smart energy system, improve the operating efficiency and reliability of the smart energy system, attract different user groups to provide related services, and thus improve the economy and safety of the power system.

[0004] The technical solution adopted by the present invention is as follows: an optimization and coordination method for a multi-layer heterogeneous smart energy system, comprising the following steps:

[0005] S1: Construct an upper-level optimization model, and construct the objective function of the upper-level optimization model with the total operating cost as the minimum target, as shown in formula (1):

[0006]

[0007]

[0008] In the formula, is the generator cost function; is the generator output power; The amount of electricity purchased from the wholesale market; To provide backup power; To provide backup power; is the price of electricity purchased from the wholesale market; The price of upper reserve electricity; is the reserve power price; T is the total duration; t is the time variable; N g is the number of generators; k is the variable of the number of generators; N re is the total number of renewable energy sources; m is the variable of the number of renewable energy sources; L m,t is the reserve loss; c u and c l are the renewable energy spillover and load shedding cost coefficients, respectively; are the upper and lower limits of the renewable energy forecast error range, respectively; is the renewable energy forecast error; are the downside reserve and upside reserve shortage values ​​respectively; P r () is the probability distribution function of renewable energy prediction error, which is obtained from historical data and obeys Gaussian distribution; ω is the risk tolerance coefficient, which indicates the tolerance of users of smart energy system to the occurrence of risks. The smaller the value, the more acceptable the risk is, and the value range is [0,1]; δ j represents the conditional risk value of user j; N c is the total number of users; N v is the total load, is the reward for the load in the i-th smart energy system, that is, the reward for the load to provide peak-shaving and valley-filling services. In order to obtain better peak-shaving and valley-filling performance, the distribution network operator releases the reference power and reward price to the smart energy system in advance to encourage the smart energy system to participate in load shaping, which is expressed by formula (3):

[0009]

[0010] In the formula, θ 1 ,θ 2 are the reward coefficients for reducing power deviation and power climbing respectively; is the reference signal of the i-th smart energy system, which is determined based on the load, renewable energy output and historical daily average grid load, with the aim of smoothing the net load curve of the distribution network operator; is the reference ramp power of the smart energy system, equal to In order to reserve the active power exchanged between the smart energy system and the distribution network operator after the load is finalized, the calculation method is shown in formula (4):

[0011]

[0012] In the formula, for The climbing δ and It is the deviation threshold of power and power ramp. If the power curve of the smart energy system is within the threshold range, it will be rewarded; otherwise, it will be punished;

[0013] The constraints include the following:

[0014] 1) Power constraint: The components located on each bus b should satisfy the following node power balance equation:

[0015]

[0016] In the formula, C bc is a binary constant. If bus b is a connection node to the main grid, then C bc =1; otherwise, C bc =0; Reactive power injected into the main grid; Output active power and reactive power for the distributed power source located at bus b; is the predicted output of the new energy plant located at bus b, which is capable of fully consuming renewable energy; are the active and reactive outputs of renewable energy sources located at bus b, respectively; are the active and reactive loads of bus b respectively; p b,t ,q b,t They represent the active power and reactive power injection of bus b respectively. The decoupled linearized power flow model is adopted, and the constraint equations are as follows:

[0017]

[0018] In the formula, G bj , B bj is the admittance matrix of the system; G b ' j , B' bj is the admittance matrix without shunt elements; g bj , b bj 、p bj,t ,q bj,t Respectively represent the conductance, susceptance, active power and reactive power of branches b to j; V b,t ,θ b,t They represent the voltage amplitude and phase angle of bus b respectively;

[0019] 2) Network constraints: The voltage amplitude and power flow should be kept within the safe range. The constraints are shown in formulas (11) and (12):

[0020]

[0021]

[0022] In the formula, are the minimum and maximum voltages of bus bar b, are the maximum active power and maximum reactive power of branches b to j respectively;

[0023] 3) Distributed power output constraints: The output of each distributed power source should meet the constraints of power output range and ramp limit, as shown in formulas (13), (14), and (15):

[0024]

[0025]

[0026] In the formula, are the minimum active power and maximum active power output by the kth distributed power source respectively; are the minimum reactive power and maximum reactive power output by the kth distributed power source; r k ,rd k are the maximum rising rate and falling rate of the ramp rate of the kth distributed generation respectively;

[0027] 4) Constraints on electricity purchase and sales prices: is the upper / lower limit of the price of electricity purchased from the wholesale market, is the upper and lower limits of the reserve power price, are the upper and lower limits of the reserve power price, as shown in formula (16):

[0028]

[0029] S2: Construct the lower-level optimization model, and construct the objective function of the lower-level optimization model with the goal of maximizing the profit of the aggregator, as shown in formula (17). The objective function includes the benefits of providing upstream and downstream reserves and energy to distribution network operators, the benefits of providing peak-shaving and valley-filling services, and the generation cost of distributed power sources owned by the smart energy system agent;

[0030]

[0031] The constraints include the following:

[0032] 1) Power balance constraint: The constraint equations are shown in formulas (18) and (19):

[0033]

[0034]

[0035] In the formula, are the active and reactive powers of the transferable loads of the i-th smart energy system, respectively; is the predicted power value of the new energy station; are the active and reactive power of the electric vehicle aggregator in the i-th smart energy system; the predicted value

[0036] 2) Distributed power output constraints: The distributed power output constraints of each smart energy system are the same as those of the upper model. The constraint equations are shown in formulas (20), (21), and (22):

[0037]

[0038]

[0039] 3) Flexible load constraint: There are two types of flexible loads involved, including transferable loads and electric vehicle loads. Formula (23) limits the power of the transferable load in each time period. Formula (24) shows that the daily energy consumption must be higher than the minimum level.

[0040]

[0041] In the formula, is the predicted active and reactive power load of the ith smart energy system, are the upper and lower limits of the displacement load, respectively;

[0042] Electric vehicles are special delayable loads with arrival and departure times. The construction of electric vehicle aggregators is shown in formulas (25) and (26):

[0043]

[0044] In the formula, is the maximum charging power of the EV aggregator; t s ,t e They are the charging start and end times of the EV aggregator; is the energy demand of EV aggregators;

[0045] The relationship between active power and reactive power of flexible loads and electric vehicle aggregators is shown in formulas (27) and (28):

[0046]

[0047] In the formula, γ lf , γ ev are the power factors of the load and the EV aggregator, respectively, is the power factor angle of the load; is the power factor angle of the EV aggregator; is the power factor angle of the load ratio aggregator;

[0048] 4) Energy storage constraint: The energy storage provided by the smart energy system to the distribution network operator is provided by its own distributed power sources. The constraint equations are shown in formulas (29), (30), (31), (32), (33), and (34):

[0049]

[0050]

[0051] In the formula, The upper and lower limits of the reserve capacity in the ith smart energy system, respectively;

[0052] S3: After the upper-layer model and the lower-layer model are established respectively, the upper-layer model and the lower-layer model are reconstructed, the lower-layer model is converted into equivalent constraints of the upper-layer model, and added to the upper-layer model, and the double-layer model is converted into a single-layer model; according to the Big-M method, the nonlinear constraint part of the single-layer model is linearized to obtain a single-layer linear model, and the single-layer linear model is directly solved by a commercial solver to obtain the optimization information of the multi-layer heterogeneous smart energy system.

[0053] Furthermore, in step S3, the upper and lower layer models are reconstructed as follows: KKT conditions need to be derived, and two sets of auxiliary variables must be introduced to include absolute values and Linearize the reward term, and in the lower objective function, replace Rv ls Replace it with formula (34) and add constraint formulas (35) and (36) to the lower model.

[0054]

[0055]

[0056] In the formula, yes The dual variable of yes The dual variable of yes The dual variable of yes The dual variable of

[0057] Then, by using the KKT optimality condition, the two-level optimization problem is converted to an equivalent single-level problem, and the nonlinear constraints are linearized by the big-M method. The equivalent mixed integer linear format is shown in equations (38) and (39):

[0058]

[0059]

[0060] Where ψ is a binary variable, M is a constant, M>9999;

[0061] The same formulas (2) and (3) are converted into formulas (40) and (41) as shown below:

[0062]

[0063] In the formula, is the first half of formula (2), is the second half of formula (2), is the deviation value of the new energy station, is the reserve loss of the upper model converted to a single-layer model, C r () is the cumulative distribution function. After the above transformation, the two-layer problem is transformed into a single-layer optimization problem.

[0064] Another object of the present invention is to be achieved through an optimization and coordination device for a multi-layer heterogeneous smart energy system, comprising: a memory, a processor, and a program stored on the memory and executable on the processor, wherein when the program is executed by the processor, an optimization and coordination method for a multi-layer heterogeneous smart energy system as described above is implemented.

[0065] Another object of the present invention is to be achieved through a computer-readable storage medium, on which a program for implementing information transmission is stored, and when the program is executed by a processor, the steps of the optimization and coordination method of a multi-layer heterogeneous smart energy system as described above are implemented.

[0066] Advantages and beneficial effects of the present invention: The present invention can coordinate various types of resources to operate efficiently, economically and greenly, thereby achieving the purpose of efficient resource calculation and optimization scheduling. At the same time, the present invention improves the efficiency of coordinated operation of power aggregators, grid reliability, and flexible adjustment and response mechanisms of power systems. At the same time, the present invention can reduce the operating costs of smart energy systems and attract different user groups to provide related services, thereby improving the economy and safety of power systems. Moreover, the present invention has a far-reaching impact on promoting the integration of renewable energy, reducing greenhouse gas emissions, enhancing the competitiveness of the energy market, and improving users' energy consumption experience, and is of great significance. BRIEF DESCRIPTION OF THE DRAWINGS

[0067] Figure 1 This is a flow chart of the optimization coordination method of the present invention. DETAILED DESCRIPTION

[0068] The present invention is further described below with reference to examples in the accompanying drawings of the specification, and the present invention is described through some embodiments. It is known to those skilled in the art that various changes or equivalent substitutions may be made to these features and embodiments without departing from the spirit and scope of the present invention.

[0069] Example 1

[0070] like Figure 1 As shown, an optimization and coordination method for a multi-layer heterogeneous smart energy system includes the following steps:

[0071] S1: Constructing the upper optimization model: To consider the adequacy of the system's reserve capacity and the type of network between the system grid operator and the user group, the operation calculation is optimized to minimize the total operating cost of the distribution system, and the impact of changes in energy and reserve prices on the system is taken into account. While considering the user's risk tolerance, users are encouraged to provide auxiliary services, and the system's operating reliability is improved while reducing operating costs. Including load rewards, user risk tolerance, electricity purchase and sales prices, power, network, distributed power output model and constraints of the smart energy system, this embodiment constructs the objective function of the upper optimization model with the total operating cost as the minimum target, as shown in formula (1):

[0072]

[0073]

[0074] In the formula, is the generator cost function; is the generator output power; The amount of electricity purchased from the wholesale market; To provide backup power; To provide backup power; is the price of electricity purchased from the wholesale market; The price of upper reserve electricity; is the reserve power price; T is the total duration; t is the time variable; N g is the number of generators; k is the variable of the number of generators; N re is the total number of renewable energy sources; m is the variable of the number of renewable energy sources; L m,t is the reserve loss; c u and c l are the renewable energy spillover and load shedding cost coefficients, respectively; are the upper and lower limits of the renewable energy forecast error range, respectively; is the renewable energy forecast error; are the downside reserve and upside reserve shortage values ​​respectively; P r() is the probability distribution function of renewable energy prediction error, which is obtained from historical data and obeys Gaussian distribution; ω is the risk tolerance coefficient, which indicates the tolerance of users of smart energy system to the occurrence of risks. The smaller the value, the more acceptable the risk is, and the value range is [0,1]; δ j represents the conditional risk value of user j; N c is the total number of users; N v is the total load, is the reward for the load in the i-th smart energy system, that is, the reward for the load to provide peak-shaving and valley-filling services. For distribution network operators, reducing the peak-valley difference can improve the safety and economy of system operation. In order to obtain better peak-shaving and valley-filling performance, the distribution network operator releases the reference power and reward price to the smart energy system in advance to encourage the smart energy system to participate in load shaping, which is expressed by formula (3):

[0075]

[0076] In the formula, θ 1 ,θ 2 are the reward coefficients for reducing power deviation and power climbing respectively; is the reference signal of the i-th smart energy system, which is determined based on the load, renewable energy output and historical daily average grid load, with the aim of smoothing the net load curve of the distribution network operator; is the reference ramp power of the smart energy system, equal to In order to reserve the active power exchanged between the smart energy system and the distribution network operator after the load is finalized, the calculation method is shown in formula (4):

[0077]

[0078] In the formula, for The climbing and It is the deviation threshold of power and power ramp. If the power curve of the smart energy system is within the threshold range, it will be rewarded; otherwise, it will be punished;

[0079] The constraints include the following:

[0080] 1) Power constraint: The components located on each bus b should satisfy the following node power balance equation:

[0081]

[0082] In the formula, C bc is a binary constant. If bus b is a connection node to the main grid, then C bc =1; otherwise, C bc =0; Reactive power injected into the main grid; Output active power and reactive power for the distributed power source located at bus b; is the predicted output of the new energy plant located at bus b, which is capable of fully consuming renewable energy; are the active and reactive outputs of renewable energy sources located at bus b, respectively; are the active and reactive loads of bus b respectively; p b,t ,q b,t They represent the active power and reactive power injection of bus b respectively. The decoupled linearized power flow model is adopted, and the constraint equations are as follows:

[0083]

[0084] In the formula, G bj , B bj is the admittance matrix of the system; G′ bj , B' bj is the admittance matrix without shunt elements; g bj , b bj 、p bj,t ,q bj,t Respectively represent the conductance, susceptance, active power and reactive power of branches b to j; V b,t ,θ b,t They represent the voltage amplitude and phase angle of bus b respectively;

[0085] 2) Network constraints: The voltage amplitude and power flow should be kept within the safe range. The constraints are shown in formulas (11) and (12):

[0086]

[0087]

[0088] In the formula, are the minimum and maximum voltages of bus bar b, are the maximum active power and maximum reactive power of branches b to j respectively;

[0089] 3) Distributed power output constraints: The output of each distributed power source should meet the constraints of power output range and ramp limit, as shown in formulas (13), (14), and (15):

[0090]

[0091]

[0092] In the formula, are the minimum active power and maximum active power output by the kth distributed power source respectively; are the minimum reactive power and maximum reactive power output by the kth distributed power source; r k ,rd k are the maximum rising rate and falling rate of the ramp rate of the kth distributed generation respectively;

[0093] 4) Constraints on electricity purchase and sales prices: is the upper / lower limit of the price of electricity purchased from the wholesale market, is the upper and lower limits of the reserve power price, are the upper and lower limits of the reserve power price, as shown in formula (16):

[0094]

[0095] S2: Construct the lower-level optimization model. After receiving the price curve released by the grid operator, the user group solves the problem of maximizing its own interests based on the price obtained by optimization calculation, and feeds back the energy and reserved bids to the grid operator. It includes power balance, distributed power output, flexible load, and energy storage related models and constraints. The objective function of the lower-level optimization model is constructed with the goal of maximizing the profit of the aggregator, as shown in formula (17). The objective function includes the benefits of providing upstream and downstream reserves and energy to the distribution network operator, the benefits of providing peak-shaving and valley-filling services, and the power generation cost of the distributed power sources owned by the smart energy system agent;

[0096]

[0097] The constraints include the following:

[0098] 1) Power balance constraint: The constraint equations are shown in formulas (18) and (19):

[0099]

[0100] In the formula, are the active and reactive powers of the transferable loads of the i-th smart energy system, respectively; is the predicted power value of the new energy station; are the active and reactive power of the electric vehicle aggregator in the i-th smart energy system; the predicted value

[0101] 2) Distributed power output constraints: The distributed power output constraints of each smart energy system are the same as those of the upper model. The constraint equations are shown in formulas (20), (21), and (22):

[0102]

[0103]

[0104] 3) Flexible load constraint: There are two types of flexible loads involved, including transferable loads and electric vehicle loads. Formula (23) limits the power of the transferable load in each time period. Formula (24) shows that the daily energy consumption must be higher than the minimum level.

[0105]

[0106] In the formula, is the predicted active and reactive power load of the ith smart energy system, are the upper and lower limits of the displacement load, respectively;

[0107] Electric vehicles are special delayable loads with arrival and departure times. The construction of electric vehicle aggregators is shown in formulas (25) and (26):

[0108]

[0109] In the formula, is the maximum charging power of the EV aggregator; t s ,t e They are the charging start and end times of the EV aggregator; is the energy demand of EV aggregators;

[0110] The relationship between active power and reactive power of flexible loads and electric vehicle aggregators is shown in formulas (27) and (28):

[0111]

[0112] In the formula, γ lf , γ ev are the power factors of the load and the EV aggregator, respectively, is the power factor angle of the load; is the power factor angle of the EV aggregator; is the power factor angle of the load ratio aggregator;

[0113] 4) Energy storage constraint: The energy storage provided by the smart energy system to the distribution network operator is provided by its own distributed power sources. The constraint equations are shown in formulas (29), (30), (31), (32), (33), and (34):

[0114]

[0115]

[0116] In the formula, The upper and lower limits of the reserve capacity in the ith smart energy system, respectively;

[0117] S3: After the upper model and the lower model are established respectively, the upper and lower models are reconstructed, the lower model is converted into equivalent constraints of the upper model, and added to the upper model, and the double-layer model is converted into a single-layer model; according to the Big-M method, the nonlinear constraint part of the single-layer model is linearized to obtain a single-layer linear model, and the single-layer linear model is directly solved by the commercial solver cplex to obtain the optimization information of the multi-layer heterogeneous smart energy system. The method for reconstructing the upper and lower models is as follows: the KKT conditions need to be derived, and two sets of auxiliary variables containing absolute values ​​must be introduced and Linearize the reward term, and in the lower objective function, replace Rv ls Replace it with formula (34) and add constraint formulas (35) and (36) to the lower model.

[0118]

[0119]

[0120] In the formula, yes The dual variable of yes The dual variable of yes The dual variable of yes The dual variable of

[0121] Then, by using the KKT optimality condition, the two-level optimization problem is converted to an equivalent single-level problem, and the nonlinear constraints can be linearized by the big-M method. Its equivalent mixed integer linear format is shown in formulas (38) and (39):

[0122]

[0123] Where ψ is a binary variable, M is a constant, M>9999;

[0124] The same formulas (2) and (3) are converted into formulas (40) and (41) as shown below:

[0125]

[0126] In the formula, is the first half of formula (2), is the second half of formula (2), is the deviation value of the new energy station, is the reserve loss of the upper model converted to a single-layer model, C r () is the cumulative distribution function. After the above transformation, the two-layer problem is transformed into a single-layer optimization problem.

Claims

1. An optimization and coordination method for a multi-layer heterogeneous smart energy system, characterized in that: The method comprises the following steps: S1: Construct an upper-level optimization model, and construct the objective function of the upper-level optimization model with the total operating cost as the minimum target, as shown in formula (1): In the formula, is the generator cost function; is the generator output power; The amount of electricity purchased from the wholesale market; To provide backup power; To provide backup power; is the price of electricity purchased from the wholesale market; The price of upper reserve electricity; is the reserve power price; T is the total duration; t is the time variable; N g is the number of generators; k is the variable of the number of generators; N re is the total number of renewable energy sources; m is the variable of the number of renewable energy sources; L m,t is the reserve loss; c u and c l are the renewable energy spillover and load shedding cost coefficients, respectively; are the upper and lower limits of the renewable energy forecast error range, respectively; is the renewable energy forecast error; are the downside reserve and upside reserve shortage values ​​respectively; P r () is the probability distribution function of renewable energy prediction error, which is obtained from historical data and obeys Gaussian distribution; ω is the risk tolerance coefficient, which indicates the degree of tolerance of users in the smart energy system to the occurrence of risks. The smaller the value, the more acceptable the risk is, and the value range is [0,1]; δ j represents the conditional risk value of user j; N c is the total number of users; N v is the total load, is the reward for the load in the i-th smart energy system, that is, the reward for the load to provide peak-shaving and valley-filling services. In order to obtain better peak-shaving and valley-filling performance, the distribution network operator releases the reference power and reward price to the smart energy system in advance to encourage the smart energy system to participate in load shaping, which is expressed by formula (3): In the formula, θ1 and θ2 are the reward coefficients for reducing power deviation and power climbing respectively; is the reference signal of the i-th smart energy system, which is determined based on the load, renewable energy output and historical daily average grid load, with the aim of smoothing the net load curve of the distribution network operator; is the reference ramp power of the smart energy system, equal to In order to reserve the active power exchanged between the smart energy system and the distribution network operator after the load is finalized, the calculation method is shown in formula (4): In the formula, for The climbing and It is the deviation threshold of power and power ramp. If the power curve of the smart energy system is within the threshold range, it will be rewarded; otherwise, it will be punished; The constraints include the following: 1) Power constraint: The components located on each bus b should satisfy the following node power balance equation: In the formula, C bc is a binary constant. If bus b is a connection node to the main grid, then C bc =1; Otherwise, C bc =0; Reactive power injected into the main grid; Output active power and reactive power for the distributed power source located at bus b; is the predicted output of the new energy plant located at bus b, which is capable of fully consuming renewable energy; are the active and reactive outputs of renewable energy sources located at bus b, respectively; are the active and reactive loads of bus b respectively; p b,t ,q b,t They represent the active power and reactive power injection of bus b respectively. The decoupled linearized power flow model is adopted, and the constraint equations are as follows: p bj,t =g bj (V b,t -V j,t )-b bj (i b,t -θ j,t ) (9) q bj,t =-b bj (V b,t -V j,t )-g bj (i b,t -θ j,t ) (10) In the formula, G bj , B bj is the admittance matrix of the system; G b ' j , B' bj is the admittance matrix without shunt elements; g bj , b bj 、p bj,t ,q bj,t Respectively represent the conductance, susceptance, active power and reactive power of branches b to j; V b,t ,θ b,t They represent the voltage amplitude and phase angle of bus b respectively; 2) Network constraints: The voltage amplitude and power flow should be kept within the safe range. The constraints are shown in formulas (11) and (12): In the formula, are the minimum and maximum voltages of bus bar b, are the maximum active power and maximum reactive power of branches b to j respectively; 3) Distributed power output constraints: The output of each distributed power source should meet the constraints of power output range and ramp limit, as shown in formulas (13), (14), and (15): In the formula, are the minimum active power and maximum active power output by the kth distributed power source respectively; are the minimum reactive power and maximum reactive power output by the kth distributed power source; r k ,rd k are the maximum rising rate and falling rate of the ramp rate of the kth distributed generation respectively; 4) Constraints on electricity purchase and sales prices: is the upper / lower limit of the price of electricity purchased from the wholesale market, is the upper and lower limits of the reserve power price, are the upper and lower limits of the reserve power price, as shown in formula (16): S2: Construct the lower-level optimization model, and construct the objective function of the lower-level optimization model with the goal of maximizing the profit of the aggregator, as shown in formula (17). The objective function includes the benefits of providing upstream and downstream reserves and energy to distribution network operators, the benefits of providing peak-shaving and valley-filling services, and the generation cost of distributed power sources owned by the smart energy system agent; The constraints include the following: 1) Power balance constraint: The constraint equations are shown in formulas (18) and (19): In the formula, are the active and reactive powers of the transferable loads of the i-th smart energy system, respectively; is the predicted power value of the new energy station; are the active and reactive power of the electric vehicle aggregator in the i-th smart energy system; the predicted value 2) Distributed power output constraints: The distributed power output constraints of each smart energy system are the same as those of the upper model. The constraint equations are shown in formulas (20), (21), and (22): 3) Flexible load constraint: There are two types of flexible loads involved, including transferable loads and electric vehicle loads. Formula (23) limits the power of the transferable load in each time period. Formula (24) indicates that the daily energy consumption must be higher than the minimum level. In the formula, is the predicted active and reactive power load of the ith smart energy system, are the upper and lower limits of the displacement load, respectively; Electric vehicles are special delayable loads with arrival and departure times. The construction of electric vehicle aggregators is shown in formulas (25) and (26): In the formula, is the maximum charging power of the EV aggregator; t s ,t e They are the charging start and end times of the EV aggregator; is the energy demand of EV aggregators; The relationship between active power and reactive power of flexible loads and electric vehicle aggregators is shown in formulas (27) and (28): In the formula, γ lf , γ ev are the power factors of the load and the EV aggregator, respectively, is the power factor angle of the load; is the power factor angle of the EV aggregator; is the power factor angle of the load ratio aggregator; 4) Energy storage constraint: The energy storage provided by the smart energy system to the distribution network operator is provided by its own distributed power sources. The constraint equations are shown in formulas (29), (30), (31), (32), (33), and (34): In the formula, The upper and lower limits of the reserve capacity in the ith smart energy system, respectively; S3: After the upper-layer model and the lower-layer model are established respectively, the upper-layer model and the lower-layer model are reconstructed, the lower-layer model is converted into equivalent constraints of the upper-layer model, and added to the upper-layer model, and the double-layer model is converted into a single-layer model; according to the Big-M method, the nonlinear constraint part of the single-layer model is linearized to obtain a single-layer linear model, and the single-layer linear model is directly solved by a commercial solver to obtain the optimization information of the multi-layer heterogeneous smart energy system.

2. The optimization and coordination method of a multi-layer heterogeneous smart energy system according to claim 1 is characterized in that: Step S3 reconstructs the upper and lower models by: KKT conditions need to be derived, and two sets of auxiliary variables must be introduced to contain absolute values. and Linearize the reward term, and in the lower objective function, replace Rv ls Replace it with formula (34) and add constraint formulas (35) and (36) to the lower model. In the formula, yes The dual variable of yes The dual variable of yes The dual variable of yes The dual variable of Then, by using the KKT optimality condition, the two-level optimization problem is converted to an equivalent single-level problem, and the nonlinear constraints are linearized by the big-M method. The equivalent mixed integer linear format is shown in equations (38) and (39): Where ψ is a binary variable, M is a constant, M>9999; The same formulas (2) and (3) are converted into formulas (40) and (41) as shown below: In the formula, is the first half of formula (2), is the second half of formula (2), is the deviation value of the new energy station, L * m,t is the reserve loss of the upper model converted to a single-layer model, C r () is the cumulative distribution function. After the above transformation, the two-layer problem is transformed into a single-layer optimization problem.

3. An optimization and coordination device for a multi-layer heterogeneous smart energy system, characterized in that: include: A memory, a processor, and a program stored in the memory and executable on the processor, wherein when the program is executed by the processor, an optimization and coordination method for a multi-layer heterogeneous smart energy system as described in any one of claims 1 to 2 is implemented.

4. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores an implementation program for information transmission, and when the program is executed by the processor, the steps of the optimization and coordination method of a multi-layer heterogeneous smart energy system as described in any one of claims 1-2 are implemented.

Images