Flexible power distribution network interaction energy control method supported by reconfigurable intelligent soft switch

Abstract

The invention discloses a flexible power distribution network interaction energy control method supported by a reconfigurable intelligent soft switch, and the method comprises the following steps: 1, constructing a reconfigurable intelligent soft switch MOP constraint, and building a power flow model of a flexible power distribution system based on the reconfigurable intelligent soft switch support; 2, constructing a power distribution network operator model, and constructing a reconfigurable intelligent soft switch MOP constraint and a power grid operator constraint; step 3, constructing a production and disappearing person model, obtaining a target function of a production and disappearing person, and constructing a production and disappearing person constraint; and 4, carrying out model linearization and relaxation processing. According to the invention, the power flow of the flexible power distribution network based on the reconfigurable intelligent soft switch is accurately regulated and controlled, and the safety guarantee capability is enhanced; a source-load multi-resource refined time sequence space-time coordinated regulation and control model is provided, so that the voltage safety under large-scale distributed photovoltaic access can be guaranteed, and the economy of system operation can be improved, namely the safe and economic operation performance of the system is improved.

Description

Technical Field

[0001] The present invention belongs to the field of power system optimization and energy management, and specifically relates to a flexible distribution network interactive energy control method supported by a reconfigurable intelligent soft switch. Background Art

[0002] As producers and consumers of energy, prosumers play an important role in the energy market, and their flexibility provides new possibilities for improving the economy and reliability of the distribution system. However, the extensive participation of prosumers also leads to an increase in the complexity of multi-agent energy interactions, while intermittent demand response poses a challenge to the balance of the system.

[0003] Existing energy interaction methods mostly focus on direct control or regional market-based control. The former is prone to privacy leakage and limits the autonomy of producers and consumers, while the latter pays insufficient attention to power quality and voltage safety issues. In addition, most traditional optimization strategies only consider economic benefits, ignoring the safety and reliability of power grid operation. They cannot meet the operating requirements under a high proportion of distributed energy access and do not consider voltage regulation issues, which may affect the stability of the system. Therefore, there is an urgent need for a new energy interaction strategy that takes into account both economy and system safety. Different from the traditional iterative method solution, the present invention uses KKT conditions, duality principles, linearization methods and relaxation techniques to transform the original two-level game problem into a single-level mixed integer second-order cone programming problem, thereby improving the solution efficiency; finally, the improved IEEE-33 node distribution system is simulated and verified, and the effectiveness of the proposed strategy is proved by comparing the three schemes. Summary of the invention

[0004] The present invention aims to solve the shortcomings of the prior art in the multi-producer-consumer distribution system, and proposes a flexible distribution network interactive energy control method supported by a reconfigurable intelligent soft switch, aiming to balance the economic benefits of multiple subjects and the safety of system operation, and improve the flexibility and reliability of the distribution system. The present invention constructs a game optimization model between the distribution network operator (DNO) and the producers and consumers, and introduces power flow control technology to improve power quality, reduce system losses, and achieve efficient energy interaction among multiple subjects.

[0005] In order to achieve the above-mentioned purpose, the present invention adopts the following technical scheme:

[0006] In one aspect of the present invention, the present invention proposes a flexible distribution network interactive energy control method supported by a reconfigurable intelligent soft switch. According to an embodiment of the present invention, the method comprises the following steps:

[0007] Step 1: Construct the MOP constraint of the reconfigurable intelligent soft switch and establish the power flow model of the flexible distribution system based on the reconfigurable intelligent soft switch.

[0008] Step 2: Based on step 1, build a distribution network operator model

[0009] Step 2.1 constructs the MOP constraint of the reconfigurable intelligent soft switch and the power flow model of the flexible distribution system supported by the reconfigurable intelligent soft switch, thereby obtaining the objective function of the distribution network operator;

[0010] Step 2.2 constructs grid operator constraints, including DN security constraints and price constraints;

[0011] Step 3: Build a producer-consumer model

[0012] Step 3.1 obtain the objective function of the producer and consumer;

[0013] Step 3.2 constructs constraints for prosumers, including energy storage ESS operation constraints and intermittent demand response DR constraints;

[0014] Step 4: Model linearization and relaxation

[0015] Step 4.1 Linearization processing, including quadratic term processing and absolute value term processing;

[0016] Step 4.2: Second-order cone transformation;

[0017] Step 4.3 Two-layer model and its transformation.

[0018] In addition, the flexible distribution network interactive energy control method supported by the reconfigurable intelligent soft switch according to the above embodiment of the present invention may also have the following additional technical features:

[0019] In some embodiments of the present invention, in step 1, the reconfigurable intelligent soft switch MOP constraint is constructed as follows:

[0020]

[0021]

[0022]

[0023]

[0024]

[0025]

[0026]

[0027] Among them, equations (1) to (3) are power balance constraints, Ω VSC A collection of feeders connected to the VSC; is the active power of the DC side voltage source converter VSC at node i at time t; is VSC at time t i The actual transmitted active power; is the MOP at time t i Active power loss; For VSC i The loss coefficient of m is the total number of branches connected to the MOP; is the active power of VSC on the DC side at time t; is the capacity of the VSC connected to node i;

[0028] Formula (4) to (7) are MOP capacity constraints, where: It represents the power transmission capacity of the nth branch connected to the MOP; Represents VSC i Capacity; is VSC at time t i The reactive power actually transmitted; B i,n Indicates the switch status of the MOP connected to node i and the nth branch connected to it; For VSC i Reactive power output limit; N It is the branch connected to MOP.

[0029] In some embodiments of the present invention, in step 1, a power flow model of a flexible power distribution system supported by a reconfigurable intelligent soft switch is constructed as follows:

[0030]

[0031]

[0032]

[0033]

[0034]

[0035]

[0036] Among them, equations (8) and (9) show the active and reactive power balance of the branch, P t,ji , Q t,ji are the active and reactive power flowing through branch ij at time t, P t,i , Q t,i is the active and reactive power injected into node i at time t; r ji 、x ij are the resistance and reactance of the branch between node j and node i respectively; is the square of the current in the branch between node j and node i at time t; Ω b is the set of branches connected to VSC; P t,ik , Q t,ik are the active and reactive power flowing through branch ik at time t respectively; Equations (10) and (11) are the constraints on node voltage and branch current levels at time t; are the squares of the voltages on nodes i and j at time t; r ij 、x ij are the resistance and reactance of the branch between node i and node j respectively; P t,ij , Q t,ij are the active and reactive power flowing through branch ij at time t respectively; is the square of the current on the branch between node i and node j at time t; Equations (12) and (13) indicate the active and reactive balance of the node. are the active power injected into node i by PV and VSC at time t respectively; are the fixed load and transferable load of the producer and consumer at node i at time t, A positive value indicates an increase in the load demand of node i at time t, and a negative value indicates a decrease in the load demand of node i at time t; are the charging and discharging power of ESS to node i at time t respectively; is the reactive power of the prosumer at node i at time t; is the reactive power injected into node i by VSC at time t;

[0037] In some embodiments of the present invention, in step 2.1, the objective function of the distribution network operator is obtained as follows:

[0038] f DNO =a o (f Grid +f loss -f inc )+b v f vd (14)

[0039]

[0040] Formula (14) is the general objective function of the prosumer-distribution network operator DNO optimization problem, which is a linear weighted combination of the operating cost and voltage deviation minimization problem; where a o With b v is the weight coefficient, which means the attention paid to each factor in the target, and the sum is 1; f Grid With f inc represents the transaction cost between DNO and the high voltage grid HV, and the total revenue from transactions between DNO and each producer and consumer; floss is the network loss cost of the distribution network DN; f vd is the total voltage deviation; F DNO is the objective function of the grid operator;

[0041] Formula (15) is the specific expression of each element, where X t , W t are the electricity purchase and sale prices of DN and HV at time t; g t is the net load of the DN system at time t, g t If it is positive, it means that DN buys electric energy from HV at time t, g t If it is negative, it means that DN sells electric energy to HV at time t; C loss represents the cost coefficient related to network loss; r ij 、x ij is the resistance and reactance of branch ij; I t.ij is the current of branch ij in period t; is the active power loss generated by MOP in period t; N T 、N N 、N P are the total time period, total number of nodes and total number of producers and consumers in DN respectively; Δt is the time interval; Ω l is the set of all lines; t is the clearing price of the internal energy trading market in period t; represents the net load of the nth prosumer in period t, which is also the exchange power between it and DN; U t.i is the voltage of node i in period t; is the node voltage reference value; f vd It shows the voltage of each node of DN during the whole operation period. Relative to reference value The voltage deviation is used to evaluate the voltage quality of the system.

[0042] In some embodiments of the present invention, in step 2.2:

[0043] The security constraints for constructing DN are as follows:

[0044]

[0045]

[0046] In formula (16)-formula (17) U , are the upper and lower limits of the node voltage respectively; is the maximum current limit of the branch;

[0047] Construct the price constraint as follows:

[0048]

[0049]

[0050] Formula (18) shows that λ t Cannot exceed upper and lower bounds and Formula (19) shows that the mean of the clearing price is not higher than the mean of the price of electricity purchased from HV; T represents the time of 24 hours;

[0051] In some embodiments of the present invention, in step 3.1, the objective function of the prosumer is as follows:

[0052]

[0053]

[0054] Formula (20) is the general function of the optimization objective of the prosumer. The first term is the cost of participating in the electricity transaction, the second term is the utility cost of adjusting the load, and the third term is the loss cost of ESS operation. n is the objective function of the prosumer; is the transferable load of the prosumer at node i at time t, A positive value indicates an increase in the load demand of node i at time t, and a negative value indicates a decrease in the load demand of node i at time t; represents the discomfort sensitivity coefficient of the nth prosumer to load changes; is the exchange power between each prosumer and DNO at time t; Δt is the change in energy interaction time; is the active power charged on node i at time t; η ess,c , η ess,d Respectively represent the ESS charging and discharging efficiency; l deg represents the loss coefficient of ESS; Formula (21) is the exchange power function, N e represents the number of ESSs connected by prosumers;

[0055] In some embodiments of the present invention, in step 3.2:

[0056] The operating constraints for building an energy storage ESS are as follows:

[0057]

[0058]

[0059]

[0060]

[0061] Formula (22) is the ESS charging and discharging power constraint, is the working state of the e-th ESS at time t, which is 1 when charging and 0 when discharging; is the maximum value of charging and discharging power of the e-th ESS;

[0062] Formula (23)-(24) are the ESS energy storage constraints, where S t,e is the current energy storage capacity of the e-th ESS at time t, S t-1,e is the energy storage at the previous moment, and there is a recursive relationship between the two; are the upper and lower bounds of the energy storage capacity of the e-th ESS; is the maximum capacity of the e-th ESS;

[0063] Formula (25) represents the initial energy storage capacity S of the e-th ESS 1,e And last minute energy storage equal;

[0064] The intermittent demand response DR constraints are constructed as follows:

[0065]

[0066]

[0067] Formula (26) shows that the transferable load adjustment of the producer and consumer cannot exceed the range [L min ,L max ]; L min is the minimum value of the transfer load adjustment, L min is the maximum value of the transfer load regulation; Formula (27) shows that the total load demand in a day remains unchanged.

[0068] In some embodiments of the present invention, the linearization process in step 4.1 is as follows:

[0069] (1) Secondary term processing

[0070] Since equations (2), (3), (4), (5), (6), (9), and (10) have nonlinear forms such as the square of current and voltage, we use v t,i and l t,ij Replace and The converted functions are shown in equations (29) to (34):

[0071]

[0072]

[0073]

[0074]

[0075]

[0076]

[0077]

[0078] (2) Absolute value item processing

[0079] Since there is an absolute value term of voltage deviation in equation (2), an auxiliary variable Aux is introduced: t,i After linearization, the transformed function is shown in the following equations (35) to (38):

[0080]

[0081] Aux t,i ≥0(36)

[0082]

[0083]

[0084] In some embodiments of the present invention, in step 4.2, the second-order cone transformation is as follows:

[0085] The convex relaxation of formula (32) is as follows:

[0086]

[0087] The rotation second-order cone constraint after transformation of equation (14) is:

[0088]

[0089] In order to verify the relaxation effect, the following function is defined:

[0090]

[0091] If the gap value is small enough, the accuracy after relaxation is considered reasonable;

[0092] In some embodiments of the present invention, in step 4.3, the two-layer model and its transformation are as follows:

[0093] Based on the Stackelberg master-slave game, the following two-layer model is established:

[0094] Upper level goal: minF DNO ;

[0095] Upper-level constraints: equation (7)-equation (8), equation (11)-equation (13), equation (15)-equation (19), equation (29)-equation (31), equation (33)-equation (34), equation (36)-(40);

[0096] Variable: λ t , l t,ij ,v t,i ,Aux t,i ;

[0097] Lower level target: minJ n ;

[0098] Lower layer constraints: Equation (22)-Equation (27);

[0099] variable: S t,e ;

[0100] The above two-layer model is transformed using the KKT condition. Let μ be the dual variable of the inequality constraint of the lower optimization problem, and λ be the dual variable of the equality constraint of the lower optimization problem. The general formula of the converted KKT condition is:

[0101]

[0102] 0≤μ⊥g(x)≥0 (43)

[0103] is the Lagrangian function written under KKT conditions, g(x)≥0 represents the inequality constraint in the optimization problem, and the specific expression is as follows:

[0104] (1) Introducing the objective function and constraints of the lower-level prosumers to write the Lagrangian function

[0105]

[0106] (2) Taking partial derivatives of ▽L, we obtain the equality constraint

[0107] Introduce the objective function and constraints of the lower-level prosumers and write the Lagrangian function formula:

[0108]

[0109]

[0110]

[0111]

[0112]

[0113]

[0114]

[0115] (3) Using the Big M method to construct inequality constraints

[0116] Since the complementary slack variables in the KKT condition exist in the form of μ i g i (x) = 0, so we introduce a Boolean variable ε and a positive number M to construct the following linear inequality:

[0117]

[0118]

[0119]

[0120]

[0121]

[0122]

[0123]

[0124] (4) Single-layer optimization model

[0125] Substituting equation (45)-equation (51) into equation (44), we can obtain:

[0126]

[0127] Formula (60) is combined with the upper-layer DNO objective function formula (1) to obtain the objective function of the single-layer optimization model:

[0128]

[0129] Constraints: formula (7)-formula (8), formula (11)-formula (13), formula (15)-formula (19), formula (29)-formula (31), formula (33)-formula (34), formula (36)-formula (40), formula (45)-formula (59);

[0130] Decision variable: λ t , S t,e .

[0131] Compared with the prior art, the present invention has the following beneficial effects:

[0132] 1. This invention proposes a two-layer optimization strategy based on Stackelberg game. The upper layer is DNO to maximize social welfare through clearing prices, and the lower layer producers and consumers dynamically adjust energy strategies according to prices. This method takes into account both economy and safety, and solves the problem of single-objective optimization in existing strategies.

[0133] 2. The present invention introduces MOP to adjust the active and reactive power distribution of the feeder, significantly reducing power loss and voltage deviation and optimizing power quality. Compared with traditional voltage control methods, MOP has higher response speed and flexibility, effectively avoiding the problem of voltage exceeding the limit during system operation.

[0134] 3. The present invention adopts KKT conditions and linearization technology to transform the two-layer game model into a single-layer mixed integer second-order cone programming problem, which significantly improves the solution efficiency compared with the traditional iterative solution. In addition, the simplified solution process avoids unnecessary information interaction and has unique advantages in protecting the privacy of producers and consumers.

[0135] 4. The present invention can reduce the electricity cost of multiple producers and consumers, while increasing the revenue of DNO. By optimizing the power flow distribution and resource allocation, the operation safety of the system is guaranteed, and the local consumption of distributed energy is further supported, thereby improving the overall energy utilization rate and sustainability. BRIEF DESCRIPTION OF THE DRAWINGS

[0136] Figure 1 It is a flow chart of a method for interactive energy control of a flexible distribution network supported by a reconfigurable intelligent soft switch in an embodiment of the present invention;

[0137] Figure 2 The power output of the MOP in the application example of the present invention, (a) active power output; (b) reactive power output;

[0138] Figure 3 The initial data of each prosumer in the application example of the present invention are: (a) photovoltaic output; (b) load demand;

[0139] Figure 4 The comparison of voltage per unit values ​​in the application examples of the present invention, (a) comparison of maximum voltage values; (b) comparison of minimum voltage values ​​and voltage per unit values;

[0140] Figure 5 This is the improved IEEE-33 node system diagram in the application example of the present invention. DETAILED DESCRIPTION

[0141] 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.

[0142] Example

[0143] like Figure 1 As shown in the figure, the interactive energy control method of the flexible distribution network supported by the reconfigurable intelligent soft switch is firstly constructed, and the power flow model of the flexible distribution system supported by the reconfigurable intelligent soft switch is established. On the basis of constructing the MOP constraint of the reconfigurable intelligent soft switch and the power flow model of the flexible distribution system supported by the reconfigurable intelligent soft switch, the objective function of the distribution network operator is obtained, and the security constraint and price constraint of DN are considered to obtain the grid operator constraint. Secondly, the objective function of the prosumer is obtained, and the prosumer model is obtained by considering the operation constraint of the energy storage ESS and the intermittent demand response DR constraint. Finally, the model is linearized, including the quadratic term processing and the absolute value term processing, and the second-order cone transformation is performed to establish a double-layer model and transform it.

[0144] Specifically, the method comprises the following steps:

[0145] Step 1: Model the reconfigurable intelligent soft switch and establish a power flow model of a flexible distribution system based on the reconfigurable intelligent soft switch, including the following steps:

[0146] 1.1. Constructing the reconfigurable intelligent soft switch MOP constraint from equations (1) to (7)

[0147]

[0148]

[0149]

[0150]

[0151]

[0152]

[0153]

[0154] Formula (1) to (3) are power balance constraints, where Ω VSC A collection of feeders connected to the VSC; is the active power of the DC side voltage source converter VSC at node i at time t; is VSC at time t i The actual transmitted active power; is the MOP at time t i Active power loss; For VSC i The loss factor of VSC is the VSC set. Equations (4) to (7) are MOP capacity constraints, where: It represents the power transmission capacity of the nth branch connected to the MOP; Represents VSC i Capacity; is VSC at time t i The reactive power actually transmitted; B i,n Indicates the switch status of the MOP connected to node i and the nth branch connected to it; For VSC i Reactive power output limit; N m is the total number of branches connected to the MOP; is the active power of VSC on the DC side at time t; is the capacity of the VSC connected to node i, and N is the branch connected to the MOP.

[0155] 1.2. The power flow model of the flexible distribution system supported by reconfigurable intelligent soft switches is constructed by equations (8) to (13):

[0156]

[0157]

[0158]

[0159]

[0160]

[0161]

[0162] Equations (8) and (9) show that the active and reactive power balance of the branch, P t,ji , Q t,ji are the active and reactive power flowing through branch ij at time t, P t,i , Q t,i is the active and reactive power injected into node i at time t; r ji 、x ij are the resistance and reactance of the branch between node j and node i respectively; is the square of the current in the branch between node j and node i at time t; Ω b is the set of branches connected to VSC; P t,ik , Q t,ik are the active and reactive power flowing through branch ik at time t, respectively; Equations (10) and (11) are the constraints on node voltage and branch current levels at time t, are the squares of the voltages on nodes i and j at time t; r ij 、x ij are the resistance and reactance of the branch between node i and node j respectively; P t,ij , Q t,ij are the active and reactive power flowing through branch ij at time t respectively; is the square of the current on the branch between node i and node j at time t; Equations (12) and (13) indicate the active and reactive balance of the node, where: are the active and reactive powers injected into node i by PV and MOP at time t, respectively; are the fixed load and transferable load of the producer and consumer at node i at time t, A positive value indicates an increase in the load demand of node i at time t, and a negative value indicates a decrease in the load demand of node i at time t; are the charging and discharging power of ESS to node i at time t respectively; is the reactive power of the prosumer at node i at time t; is the reactive power injected into node i by the VSC at time t.

[0163] Step 2: Build a distribution network operator model

[0164] 2.1. The objective function of the distribution network operator is obtained from equations (14) and (15):

[0165] f DNO =a o (f Grid +f loss -f inc )+b v f vd (14)

[0166]

[0167] Formula (14) is the general objective function of the prosumer-distribution network operator DNO optimization problem, which is a linear weighted combination of the operating cost and voltage deviation minimization problem. o With b v is the weight coefficient, which means the attention paid to each factor in the target, and the sum is 1; f Grid With f increpresents the transaction cost between DNO and the high voltage grid HV, and the total revenue from transactions between DNO and each producer and consumer; f loss is the network loss cost of the distribution network DN; f vd is the total voltage deviation; F DNO is the objective function of the power grid operator. Formula (15) is the specific expression of each element, where X t , W t are the electricity purchase and sale prices of DN and HV at time t; g t is the net load of the DN system at time t, g t If it is positive, it means that DN buys electric energy from HV at time t, g t If it is negative, it means that DN sells electric energy to HV at time t; C loss represents the cost coefficient related to network loss; r ij 、x ij is the resistance and reactance of branch ij; I t.ij is the current of branch ij in period t; is the active power loss generated by MOP in period t; N T 、N N 、N P are the total time period, total number of nodes and total number of producers and consumers in DN respectively; Δt is the time interval; Ω l is the set of all lines; t is the clearing price of the internal energy trading market in period t; It represents the net load of the nth prosumer in period t, and also the exchange power between it and DN. t.i is the voltage of node i in period t; is the node voltage reference value; f vd It shows the voltage of each node of DN during the whole operation period. Relative to reference value Voltage deviation is used to evaluate the voltage quality of the system;

[0168] 2.2. Constructing grid operator constraints

[0169] 2.2.1. Construct the security constraints of DN from equations (16) and (17):

[0170]

[0171]

[0172] In formula (16)-formula (17) U , are the upper and lower limits of the node voltage respectively; is the maximum current limit of the branch;

[0173] 2.2.2. Constructing price constraints from equations (18) and (19)

[0174]

[0175]

[0176] Formula (18) shows that λ t Cannot exceed upper and lower bounds and Avoid consumers not purchasing electricity due to overly high prices, which will affect the normal life of users. Formula (19) shows that the average clearing price is not higher than the average price of purchasing electricity from HV to protect the interests of users; T represents the time of 24 hours;

[0177] Step 3: Build a producer-consumer model

[0178] 3.1. The objective function of the producer and consumer is obtained from equations (21) and (22):

[0179]

[0180]

[0181] Formula (20) is the general function of the optimization objective of the prosumer. The first term is the cost of participating in the electricity transaction, the second term is the utility cost of adjusting the load, and the third term is the loss cost of ESS operation. n is the objective function of the prosumer; is the transferable load of the prosumer at node i at time t, A positive value indicates an increase in the load demand of node i at time t, and a negative value indicates a decrease in the load demand of node i at time t; represents the discomfort sensitivity coefficient of the nth prosumer to load changes; is the exchange power between each prosumer and DNO at time t; Δt is the change in energy interaction time; is the active power charged on node i at time t; η ess,c , η ess,d Respectively represent the ESS charging and discharging efficiency; l deg represents the loss coefficient of ESS; Formula (21) is the exchange power function, N e Represents the number of ESSs connected by prosumers.

[0182] 3.2. Constructing constraints on producers and consumers:

[0183] 3.2.1. Construct the energy storage ESS operation constraints from equations (22) to (25):

[0184]

[0185]

[0186]

[0187]

[0188] Formula (22) is the ESS charging and discharging power constraint, is the working state of the e-th ESS at time t, which is 1 when charging and 0 when discharging; is the maximum charging and discharging power of the e-th ESS. Equations (23) and (24) are the ESS energy storage constraints, where S t,e is the current energy storage capacity of the e-th ESS at time t, S t-1,e is the energy storage at the previous moment, and there is a recursive relationship between the two; are the upper and lower bounds of the energy storage capacity of the e-th ESS; is the maximum capacity of the e-th ESS; Formula (25) represents the initial energy storage S of the e-th ESS 1,e And last minute energy storage equal;

[0189] 3.2.2. Construct intermittent demand response DR constraints using equations (26) and (27):

[0190]

[0191]

[0192] Formula (26) shows that the transferable load adjustment of the producer and consumer cannot exceed the range [L min ,L max ]; L min is the minimum value of the transfer load adjustment, L max is the maximum value of the transfer load regulation; Formula (27) shows that the total load demand in a day remains unchanged;

[0193] Step 4: Model linearization and relaxation

[0194] 4.1 Linearization processing:

[0195] There are a lot of nonlinear functions in the objective function and constraints in the above model, but existing commercial solvers cannot solve them. Therefore, in this section, the linearization method and second-order cone relaxation technology are used to process the model and transform it into a mixed integer second-order cone programming model;

[0196] 4.1.1. Secondary term processing

[0197] Since equations (2), (3), (4), (5), (6), (9), and (10) have nonlinear forms such as the square of current and voltage, we use v t,i and l t,ij Replace and The converted functions are shown in equations (29) to (34):

[0198]

[0199]

[0200]

[0201]

[0202]

[0203]

[0204]

[0205] 4.1.2 Absolute value item processing

[0206] Since there is an absolute value term of voltage deviation in equation (2), an auxiliary variable Aux is introduced: t,i After linearization, the transformed function is shown in the following equations (35) to (38):

[0207]

[0208] Aux t,i ≥0 (36)

[0209]

[0210]

[0211] 4.2 Second-order cone transformation

[0212] Even though equation (32) has been linearized, it still has the form t,ij v t,i The quadratic nonlinear term needs to be processed again, and the form after convex relaxation is as follows:

[0213]

[0214] Similarly, there is a quadratic nonlinear term in equation (14), and the transformed rotation second-order cone constraint is:

[0215]

[0216] In order to verify the relaxation effect, the following function is defined:

[0217]

[0218] If the gap value is small enough, the accuracy after relaxation is considered reasonable, which also means that the initial model can be transformed into a model that can be solved by a commercial solver through the above two processing methods;

[0219] 4.3 Two-layer model and its transformation

[0220] Based on the Stackelberg master-slave game, combined with the above framework and model, the following two-level optimization problem is established:

[0221] Upper level goal: minF DNO ;

[0222] Upper-level constraints: equation (7)-equation (8), equation (11)-equation (13), equation (15)-equation (19), equation (29)-equation (31), equation (33)-equation (34), equation (36)-(40);

[0223] variable:

[0224] Lower level target: minJ n ;

[0225] Lower layer constraints: Equation (22)-Equation (27);

[0226] variable: S t,e

[0227] For this two-level optimization model, the traditional iterative solution is used, that is, DNO determines the clearing price λ t , the prosumer according to λ t Develop electricity consumption strategy and determine the power exchange with DNO At the same time, when writing the program, it is necessary to ensure that the constraints of the upper and lower layers can be satisfied and matched at the same time, which will significantly increase the difficulty of solving the problem. In addition, considering the privacy and security of users and avoiding excessive unnecessary information interaction between prosumers and DNOs, the present invention uses KKT conditions to transform the above two-layer model to further improve the computing efficiency and protect user privacy;

[0228] Let μ be the dual variable of the inequality constraint of the lower optimization problem, and λ be the dual variable of the equality constraint of the lower optimization problem. The general formula of the KKT condition obtained by transformation is:

[0229]

[0230] 0≤μ⊥g(x)≥0 (43)

[0231] is the Lagrangian function written under KKT conditions, g(x)≥0 represents the inequality constraint in the optimization problem, and the specific expression is as follows.

[0232] (1) Introduce the objective function and constraints of the lower-level producers and consumers and write the Lagrangian function

[0233]

[0234] (2) Taking partial derivatives of ▽L, we obtain the equality constraint

[0235] Introduce the objective function and constraints of the lower-level prosumers and write the Lagrangian function formula:

[0236]

[0237]

[0238]

[0239]

[0240]

[0241]

[0242]

[0243] (3) Using the Big M method to construct inequality constraints

[0244] Since the complementary slack variables in the KKT condition exist in the form of μ i g i (x) = 0, so we introduce the Boolean variable ε and a very large positive number M to construct the following linear inequality:

[0245]

[0246]

[0247]

[0248]

[0249]

[0250]

[0251]

[0252] (4) Single-layer optimization model

[0253] Substituting equation (45)-equation (51) into equation (44), we can obtain:

[0254]

[0255] Formula (60) is combined with the upper-layer DNO objective function formula (1) to obtain the objective function of the single-layer optimization model:

[0256]

[0257] Constraints: formula (7)-formula (8), formula (11)-formula (13), formula (15)-formula (19), formula (29)-formula (31), formula (33)-formula (34), formula (36)-formula (40), formula (45)-formula (59);

[0258] Decision variable: λ t , S t,e .

[0259] Through the above method, the two-layer game optimization problem has been transformed into a single-layer optimization problem.

[0260] Verification Example

[0261] In order to verify the effectiveness of the method in Example 1, the present invention uses Figure 5 The example analysis of the system shown in the figure is carried out. The example includes 5 prosumers connected to PV, of which the prosumers with 10 nodes and 29 nodes are equipped with ESS. The entire DN contains five PVs, two ESSs and a four-feeder MOP. The relevant parameters in the test and analysis process are shown in Table 1, and the parameters of other node branches are standard IEEE-33 bus systems. The DNO price is set to 400￥ / MWh, and the reactive power impact of renewable energy is not considered. Other parameter settings are shown in Table 2. The photovoltaic output and load demand forecasts of each prosumer are shown in Table 2. Figure 3 .

[0262] In order to fully demonstrate the effectiveness of the energy interaction strategy of the multi-producer-consumer distribution system based on game theory, the example part sets up three scenarios 1, 2 and 3 for comparison. Through simulation tests of the three scenarios, the test results are compared as shown in Table 3, including the system power loss cost (f loss ) including line power loss and MOP power loss, voltage deviation conversion cost (f vd ), DNO total operating cost (F DNO ) and the total operating cost of the prosumer (J n ).

[0263] Scenario 1: Only the system economy is considered, without considering safety issues such as MOP power flow optimization, power loss and voltage deviation.

[0264] Scenario 2: The method proposed in the embodiment of the present invention does not consider the adjustment of MOP.

[0265] Scenario 3: The method proposed in the embodiment of the present invention.

[0266] The simulation program was implemented in the Matlab environment in a computer with Windows 10, Intel(R) CoreTM i5 CPU@3.5GHz, and 8GB memory.

[0267] Table 1 Relevant parameters during the test analysis

[0268]

[0269]

[0270] Table 2 Other parameters during the test analysis

[0271]

[0272] Table 3 Simulation test results of three scenarios

[0273]

[0274] As can be seen from the above table, the total daily operating cost of DNO in scenario 1, which only considers economic benefits, is 13.95 yuan, which is about 13 times less than 193.04 yuan in scenario 2. However, the voltage deviation conversion cost increases from 2.96 yuan to 104.21 yuan, which is about 34 times more. This shows that although the scenario without considering safety performance can reduce the system operation cost, the system will adjust all controllable resources to maximize the benefits, but the power grid fluctuates greatly, the power quality is poor, there are great safety hazards, and it is not conducive to the reliable operation of the system.

[0275] Compared with Scenario 2, Scenario 3 introduces MOP. Although the power loss of MOP increases by 30.94 yuan, the network loss and voltage deviation costs are reduced. The total operating cost of DNO is reduced from 193.04 yuan to 86.72 yuan, a reduction of about 1.2 times, and the economic benefits and operational safety are significantly improved.

[0276] Figure 4 Comparison of the maximum and minimum voltage values ​​in three scenarios. Figure 2The active and reactive power output of MOP. The active power regulation of MOP is consistent with the energy storage action and transferable load action of each producer and consumer, which shows that MOP has the ability to quickly and accurately adjust the active power flow distribution of DN, avoiding transaction delays caused by line congestion during peak power consumption periods, indirectly improving the economy of the system, and can alleviate voltage fluctuations, avoid voltage over-limit, and improve the stability and safety of the system.

[0277] The above contents are merely examples and explanations of the structure of the present invention. The technicians in this technical field may make various modifications or additions to the specific embodiments described or replace them in a similar manner. As long as they do not deviate from the structure of the present invention or exceed the scope defined by the claims, they should all fall within the protection scope of the present invention.

Claims

1. A flexible distribution network interactive energy control method supported by a reconfigurable intelligent soft switch, characterized in that: The following steps are involved: Step 1: Construct the MOP constraint of the reconfigurable intelligent soft switch and establish the power flow model of the flexible distribution system based on the reconfigurable intelligent soft switch; Step 2: Based on step 1, construct a distribution network operator model; Step 2.1 constructs the MOP constraint of the reconfigurable intelligent soft switch and the power flow model of the flexible distribution system supported by the reconfigurable intelligent soft switch, thereby obtaining the objective function of the distribution network operator; Step 2.2 constructs grid operator constraints, including DN security constraints and price constraints; Step 3: Build a producer-consumer model Step 3.1: Obtain the objective function of the producer and consumer; Step 3.2 constructs constraints for prosumers, including energy storage ESS operation constraints and intermittent demand response DR constraints; Step 4: Model linearization and relaxation Step 4.1 Linearization processing, including quadratic term processing and absolute value term processing; Step 4.2: Second-order cone transformation; Step 4.3 Two-layer model and its transformation.

2. The interactive energy control method for a flexible distribution network supported by a reconfigurable intelligent soft switch according to claim 1 is characterized in that: In step 1, the reconfigurable intelligent soft switch MOP constraint is constructed as follows: Among them, equations (1) to (3) are power balance constraints, Ω VSC A collection of feeders connected to the VSC; is the active power of the DC side voltage source converter VSC at node i at time t; is VSC at time t i The actual transmitted active power; is the MOP at time t i Active power loss; For VSC i The loss coefficient of m is the total number of branches connected to the MOP; is the active power of VSC on the DC side at time t; is the capacity of the VSC connected to node i; Formula (4) to (7) are MOP capacity constraints, where: It represents the power transmission capacity of the nth branch connected to the MOP; Represents VSC i Capacity; is VSC at time t i The reactive power actually transmitted; B i,n Indicates the switch status of the MOP connected to node i and the nth branch connected to it; For VSC i Reactive power output limit; N It is the branch connected to MOP.

3. The interactive energy control method for a flexible distribution network supported by a reconfigurable intelligent soft switch according to claim 2 is characterized in that: In step 1, the power flow model of the flexible power distribution system supported by the reconfigurable intelligent soft switch is constructed as follows: Among them, equations (8) and (9) show the active and reactive power balance of the branch, P t,ji , Q t,ji are the active and reactive power flowing through branch ij at time t, P t,i , Q t,i is the active and reactive power injected into node i at time t; r ji 、x ij are the resistance and reactance of the branch between node j and node i respectively; is the square of the current in the branch between node j and node i at time t; Ω b is the set of branches connected to VSC; P t,ik , Q t,ik are respectively the active and reactive power flowing through branch ik at time t; Equations (10) and (11) are the constraints on node voltage and branch current levels at time t; are the squares of the voltages on nodes i and j at time t; r ij 、x ij are the resistance and reactance of the branch between node i and node j respectively; P t,ij , Q t,ij are the active and reactive power flowing through branch ij at time t respectively; is the square of the current in the branch between node i and node j at time t; Formula (12) and (13) show the active and reactive balance of the node. are the active power injected into node i by PV and VSC at time t respectively; are the fixed load and transferable load of the producer and consumer at node i at time t, A positive value indicates an increase in the load demand of node i at time t, and a negative value indicates a decrease in the load demand of node i at time t; are the charging and discharging power of ESS to node i at time t respectively; is the reactive power of the prosumer at node i at time t; is the reactive power injected into node i by the VSC at time t.

4. The interactive energy control method for a flexible distribution network supported by a reconfigurable intelligent soft switch according to claim 1 is characterized in that: In step 2.1, the objective function of the distribution network operator is obtained as follows: minF DNO =a o (f Grid +f loss -f inc )+b v f vd (14) Formula (14) is the general form of the objective function of the prosumer-distribution network operator DNO optimization problem, which is expressed as a linear weighted combination of the operating cost and voltage deviation minimization problems; Among them, a o With b v is the weight coefficient, which means the attention paid to each factor in the target, and the sum is 1; f Grid With f inc represents the transaction cost between DNO and the high voltage grid HV, and the total revenue from transactions between DNO and each producer and consumer; f loss is the network loss cost of the distribution network DN; f vd is the total voltage deviation; F DNO is the objective function of the grid operator; Formula (15) is the specific expression of each element, where X t , W t are the electricity purchase and sale prices of DN and HV at time t; g t is the net load of the DN system at time t, g t If it is positive, it means that DN buys electric energy from HV at time t, g t If it is negative, it means that DN sells electric energy to HV at time t; C loss represents the cost coefficient related to network loss; r ij 、x ij is the resistance and reactance of branch ij; I t.ij is the current of branch ij in period t; is the active power loss generated by MOP in period t; N T 、N N 、N P are the total time period, total number of nodes and total number of producers and consumers in DN respectively; Δt is the time interval; Ω l is the set of all lines; t is the clearing price of the internal energy trading market in period t; represents the net load of the nth prosumer in period t, which is also the exchange power between it and DN; U t.i is the voltage of node i in period t; is the node voltage reference value; f vd It shows the voltage of each node of DN during the whole operation period. Relative to reference value The voltage deviation is used to evaluate the voltage quality of the system.

5. The interactive energy control method for a flexible distribution network supported by a reconfigurable intelligent soft switch according to claim 3 is characterized in that: In step 2.2: The security constraints for constructing DN are as follows: In formula (16)-formula (17) U , are the upper and lower limits of the node voltage respectively; is the maximum current limit of the branch; Construct the price constraint as follows: Formula (18) shows that λ t Indicates the clearing price, which cannot exceed the upper and lower bounds and Formula (19) shows that the average clearing price is not higher than the average price of electricity purchased from HV; T represents the time of 24 hours.

6. The interactive energy control method for a flexible distribution network supported by a reconfigurable intelligent soft switch according to claim 5 is characterized in that: In step 3.1, the objective function of the prosumer is as follows: Formula (20) is the general function of the optimization objective of the prosumer. The first term is the cost of participating in the electricity transaction, the second term is the utility cost of adjusting the load, and the third term is the loss cost of ESS operation. n is the objective function of the prosumer; is the transferable load of the prosumer at node i at time t, A positive value indicates an increase in the load demand of node i at time t, and a negative value indicates a decrease in the load demand of node i at time t; represents the discomfort sensitivity coefficient of the nth prosumer to load changes; is the exchange power between each prosumer and DNO at time t; Δt is the change in energy interaction time; is the active power charged on node i at time t; η ess,c , η ess,d Respectively represent the ESS charging and discharging efficiency; l deg represents the loss coefficient of ESS; Formula (21) is the exchange power function, N e Represents the number of ESSs connected by prosumers.

7. The interactive energy control method for a flexible distribution network supported by a reconfigurable intelligent soft switch according to claim 6 is characterized in that: In step 3.2: The operating constraints for building an energy storage ESS are as follows: Formula (22) is the ESS charging and discharging power constraint, is the working state of the e-th ESS at time t, which is 1 when charging and 0 when discharging; is the maximum value of charging and discharging power of the e-th ESS; Formula (23)-(24) are the ESS energy storage constraints, where S t,e is the current energy storage capacity of the e-th ESS at time t, S t-1,e is the energy storage at the previous moment, and there is a recursive relationship between the two; are the upper and lower bounds of the energy storage capacity of the e-th ESS; is the maximum capacity of the e-th ESS; Formula (25) represents the initial energy storage capacity S of the e-th ESS 1,e And last minute energy storage equal; The intermittent demand response DR constraints are constructed as follows: Formula (26) shows that the transferable load adjustment of the producer and consumer cannot exceed the range [L min ,L max ]; L min is the minimum value of the transfer load regulation, L max is the maximum value of the transfer load regulation; Formula (27) shows that the total load demand in a day remains unchanged.

8. The interactive energy control method for a flexible distribution network supported by a reconfigurable intelligent soft switch according to claim 7 is characterized in that: The linearization process in step 4.1 is as follows: (1) Secondary term processing Since equations (2), (3), (4), (5), (6), (9), and (10) have nonlinear forms such as the square of current and voltage, we use v t,i and l t,ij Replace and The converted functions are shown in equations (29) to (34): (2) Absolute value item processing Since there is an absolute value term of voltage deviation in equation (2), an auxiliary variable Aux is introduced: t,i After linearization, the transformed function is shown in the following equations (35) to (38): To t,i ≥0 (36) 9. The interactive energy control method for a flexible distribution network supported by a reconfigurable intelligent soft switch according to claim 8 is characterized in that: In step 4.2, the second-order cone transformation is as follows: The convex relaxation of formula (32) is as follows: The rotation second-order cone constraint after transformation of equation (14) is: In order to verify the relaxation effect, the following function is defined: If the gap value is small enough, the accuracy after relaxation is considered reasonable.

10. The interactive energy control method for a flexible distribution network supported by a reconfigurable intelligent soft switch according to claim 9 is characterized in that: In step 4.3, the two-layer model and its transformation are as follows: Based on the Stackelberg master-slave game, the following two-layer model is established: Upper level goal: minF DNO ; Upper-level constraints: equation (7)-equation (8), equation (11)-equation (13), equation (15)-equation (19), equation (29)-equation (31), equation (33)-equation (34), equation (36)-(40); Variable: λ t , l t,ij ,v t,i ,Aux t,i ; Lower level target: minJ n ; Lower layer constraints: Equation (22)-Equation (27); variable: S t,e ; The above two-layer model is transformed using the KKT condition. Let μ be the dual variable of the inequality constraint of the lower optimization problem, and λ be the dual variable of the equality constraint of the lower optimization problem. The general formula of the converted KKT condition is: 0≤μ⊥g(x)≥0 (43) is the Lagrangian function written under KKT conditions, g(x)≥0 represents the inequality constraint in the optimization problem, and the specific expression is as follows: (1) Introduce the objective function and constraints of the lower-level producers and consumers and write the Lagrangian function (2) Taking partial derivatives of ▽L, we obtain the equality constraint Introduce the objective function and constraints of the lower-level prosumers and write the Lagrangian function formula: (3) Using the Big M method to construct inequality constraints Since the complementary slack variables in the KKT condition exist in the form of μ i g i (x) = 0, so we introduce a Boolean variable ε and a positive number M to construct the following linear inequality: (4) Single-layer optimization model Substituting equation (45)-equation (51) into equation (44), we can obtain: Formula (60) is combined with the upper-layer DNO objective function formula (1) to obtain the objective function of the single-layer optimization model: Constraints: formula (7)-formula (8), formula (11)-formula (13), formula (15)-formula (19), formula (29)-formula (31), formula (33)-formula (34), formula (36)-formula (40), formula (45)-formula (59); Decision variable: λ t , S t,e .

Images