An uncertain hierarchical dispatching method for active distribution network containing multiple microgrids
By adopting opportunity constraint planning and multihedral uncertainty ensemble in active distribution networks, and adjusting the scheduling plan using rolling optimization theory, it is solved that traditional distribution systems are difficult to access with high proportion of renewable energy. Managing the uncertainty problem of distributed power supply has achieved economic and stable power balance of the distribution network.
Patent Information
- Application Number
- CN202311350538.0
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Filing Date
- 2023-10-18
- Publication Date
- 2024-12-03
- Estimated Expiration
- 2043-10-18
AI Technical Summary
In the case where high proportion of renewable energy is connected to the active distribution network, traditional distribution systems are difficult to effectively manage the uncertainty of distributed power supplies, resulting in increased complexity of optimization scheduling of distribution networks and difficult to achieve economics in an uncertain environment power balance.
The opportunity constraint planning model is used to describe the output prediction error of renewable energy generation power generation, and the uncertainty of the microgrid interactive power plan is described through the polyhedral uncertainty set, and the output plan of the continuous scheduling equipment is adjusted using the rolling optimization theory to establish the opportunity constraints including the microgrid. The planning model, distribution network recently robust optimization scheduling model and intraday rolling optimization scheduling model are hierarchical scheduling model.
It effectively reduces the economic losses of uncertainty factors to the system, provides a more economical and feasible scheduling solution, and improves the self-balancing operation capability of the distribution network.
Smart Images

Figure CN117394366B_ABST
Abstract
Description
Technical Field
[0001] The present invention belongs to the field of distribution network operation and optimization, and specifically relates to an uncertainty hierarchical scheduling method for an active distribution network containing multiple microgrids. Background Art
[0002] With the gradual increase in the penetration rate of distributed power sources, the uncertainty factors in the distribution network have increased significantly, making the optimization and dispatching of the distribution system increasingly complex. At the same time, the transformation of the power grid to a comprehensive energy service role and the development of the power market have made the access form of distributed power sources and loads as independent entities of microgrids widely exist in the future active distribution network. The traditional distribution system optimization dispatching often only adjusts the dispatching unit in the distribution network to ensure the real-time power balance when the output of distributed power sources fluctuates greatly. However, when a high proportion of renewable energy is connected to the active distribution network in the form of microgrids, the optimized operation of dispatching units such as energy storage and controllable distributed power sources in the microgrid provides a new idea for sharing the uncertainty risk of the distribution network and improving the elasticity of the distribution network. Summary of the invention
[0003] Purpose of the invention: In order to overcome the deficiencies in the prior art, an uncertainty hierarchical scheduling strategy for active distribution networks containing multiple microgrids is provided. Based on the hierarchical scheduling framework, the uncertainty of the output of distributed renewable power sources is first described by using opportunity-constrained programming conditions, and then the uncertainty of the interactive power plan reported by the microgrid is described by using a polyhedral uncertainty set. Finally, the output plan of the continuous scheduling equipment is adjusted within the day through the rolling optimization theory, and a hierarchical scheduling model is established, which includes a microgrid day-ahead opportunity-constrained programming model, a distribution network day-ahead robust optimization scheduling model, and a distribution network intraday rolling optimization scheduling model.
[0004] Technical solution: To achieve the above-mentioned purpose, the present invention provides an uncertainty hierarchical scheduling method for an active distribution network containing multiple microgrids, comprising the following steps:
[0005] S1: Establish a day-ahead optimal dispatch model for microgrids, use the opportunity constraint programming model to describe the forecast error of renewable energy power generation output, and limit the uncertainty interval of intraday interactive power to a certain range with a certain probability level based on economic efficiency and uncertainty absorption capacity;
[0006] S2: Convert and solve the microgrid's day-ahead optimal dispatch model, and report the dispatch plan to the distribution network;
[0007] S3: Establish a day-ahead optimal dispatching model for the active distribution network, determine the action state of the discrete dispatching unit, and perform robust optimal dispatching on the distribution network side based on the interactive power uncertainty interval and load forecast information reported by the microgrid;
[0008] S4: The column and constraint generation algorithm is used to divide the robust optimization model into a main problem and subproblems with reduced variable scale;
[0009] S5: Establish an intraday optimal dispatch model for active distribution networks to improve system behavior;
[0010] S6: Use intraday rolling optimization to determine the action volume of the continuous scheduling unit within the day.
[0011] Furthermore, the expression of the day-ahead optimization dispatch model of the microgrid in step S1 is as follows:
[0012]
[0013]
[0014]
[0015]
[0016]
[0017]
[0018]
[0019]
[0020]
[0021]
[0022]
[0023] In formula (1), is the total cost of the microgrid operation on the day before; T is the total dispatching period on the day before; C t grid is the electricity purchase and sale cost between the microgrid and the upper-level distribution network during period t; C t DG is the operating cost of the distributed generation during period t; C t ESS is the operating cost of energy storage during period t; C t f is the fuel consumption cost of the gas turbine during period t; B N is a node set; and are the purchase price and sales price of electricity from the microgrid to the active distribution network in period t, respectively; is the power transmission plan of the active distribution network to the microgrid in period t; is the transmission power plan from the microgrid to the active distribution network at time t; c DG P is the operating cost of distributed power generation; t DG B is the output of distributed power generation during period t; ESS is the node set where the energy storage is located; c ESS The charging, discharging and maintenance costs of energy storage; and is the charging and discharging power of the energy storage at node i at time t. F is the node set where the gas turbine is located; c f is the fuel consumption cost per unit power of the gas turbine; is the output of the gas turbine at node i at time t; in formula (2), P t grid is the actual value of the interaction power between the active distribution network and the microgrid at time t; is the active power demand of node i during period t; in formula (3), and are the state variables of power transmission from the active distribution network to the microgrid and from the microgrid to the active distribution network at time t, respectively, which are 0-1 variables; in formula (4), Provides an upper limit on the power transmitted from the active distribution network to the microgrid; is the upper limit of the power transmitted from the microgrid to the active distribution network; in formula (5), δ is the preset interactive power prediction error range; in formula (6), The output of distributed power generation is predicted at time t; t is the output prediction error of the distributed generation at time t; in formula (7), σ t is the standard deviation of the normal distribution of the distributed generation prediction error; ρ DG and is the correlation coefficient of the prediction error; P DG is the installed capacity of distributed generation; in formula (8), and P i f are the upper and lower limits of the gas turbine output at node i respectively; in formula (9), and are the energy storage charging and discharging states at time t, respectively, and are 0-1 variables; in formula (10), are the charging and discharging power of energy storage i at time t respectively; and is the upper limit of the power value; P char and P dischar is the lower boundary of the power value; in formula (11), is the rated capacity of energy storage i; Δt is the scheduling time interval; and SOC The upper and lower limits of the state of charge of the energy storage; SOC i,tis the state of charge value of energy storage i at time t; SOC i,0 , SOC i,end , SOC i,set They are the initial state of charge value of the energy storage in the dispatching period, the state of charge value at the end of the dispatching period and the set initial state of charge value.
[0024] Furthermore, the expression for the transformation of the microgrid day-ahead optimization model in step S2 is as follows:
[0025]
[0026]
[0027]
[0028]
[0029] In formula (12), Φ -1 is the inverse function of Φ; sup is the maximum value of the sequence; in formula (13), K α To satisfy A real number.
[0030] Furthermore, the expression of the active distribution network day-ahead optimization dispatching model in step S3 is as follows:
[0031]
[0032]
[0033]
[0034]
[0035]
[0036]
[0037]
[0038]
[0039]
[0040]
[0041]
[0042]
[0043]
[0044]
[0045]
[0046]
[0047] In formula (16), Optimize the operating cost of active distribution network in the future; B N is a node set; B sub C is the node set connected to the upper grid transformer; t G is the electricity purchase cost in period t; C t LC is the dispatching cost of load reduction in period t; C t ESS is the operating cost of energy storage during period t; C t grid is the transaction price between the active distribution network and the microgrid at time t. When it is positive, it represents the revenue from the active distribution network selling electricity to the microgrid, and when it is negative, it represents the cost of the active distribution network purchasing electricity from the microgrid; λ t G P is the electricity selling price of the upper distribution network during period t; t sub B is the actual active power demand of the distribution network transformer at node i to the upper grid during period t; LC is the set of nodes where the load can be reduced; c LC Reimbursement costs for demand-side response; is the demand side response load; B ESS is the node set where the energy storage is located; c ESS The charging, discharging and maintenance costs of energy storage; and B is the charging and discharging power of energy storage at node i at time t; MG It is the set of nodes connected to the microgrid; and are the purchase price and sales price of the active distribution network for the microgrid in period t; P t gridin is the transmission power from the active distribution network to the microgrid at node i at time t; P t gridout is the transmission power from the microgrid at node i to the active distribution network at time t; in formula (17), π(j) is the set of branch headend nodes with node j as the terminal node; P j,t and Q j,t are the active injection power and reactive injection power of node j at time t respectively; P ij,t and Q ij,t are respectively the active power and reactive power of branch ij at time t; Vi,t is the node voltage of node i at time t; and are respectively the active output and reactive output of the distributed generation connected to node i at time t; and are the active load and reactive load of node i at time t respectively; is the load reduction size of node i at time t; is the reactive power output of the group switching capacitor connected to node i at time t; is the reactive power output of the static VAR compensator connected to node i at time t; tanφ is the tangent value of the power factor angle of the reduced load; in formula (18), is the square of the current of line ij at time t; is the square of the voltage of node i at time t; in formula (19), and P sub are the upper and lower limits of the upper power grid transmission power respectively; in formula (20), and are the actual daily output and the day-ahead predicted output of the microgrid connected to node i at time t; i,t ≥0, is the fluctuation rate of the distributed power supply connected to node i at time t, that is, the maximum fluctuation deviation percentage; δ i is the error range reported by the microgrid connected to node i on the previous day; τ i is the uncertain budget for the output time constraint of distributed generation, which can be used to adjust the conservatism of the solution; in formula (22), and are the energy storage charging and discharging states at time t, respectively, and are 0-1 variables; in formula (23), are the charging and discharging power of energy storage i at time t respectively; and is the upper limit of the power value; P char and P dischar is the lower boundary of the power value; in formula (24), is the rated capacity of energy storage i; Δt is the scheduling time interval; and SOC The upper and lower limits of the state of charge of the energy storage; SOC i,t is the state of charge value of energy storage i at time t; SOC i,0 , SOC i,end , SOC i,set are the initial state of charge value of the energy storage in the dispatching period, the state of charge value at the end of the dispatching period and the set initial state of charge value respectively; in formula (25), is the actual load reduction at time t; is the maximum load reduction that can be achieved at time t; is a 0-1 variable, indicating the load reduction state; in formula (26), is a 0-1 variable, indicating the start of load shedding and the stop of load shedding respectively; in formula (27), They represent the shortest load reduction time and the longest load reduction time respectively; in formula (28), represents the number of loads that can be reduced simultaneously; in formula (29), B CB is a node set including group switching capacitors; is the operational capacity of the capacitor at node i at time t, The gear switching capacity for the gear change of the group switching capacitor; is an integer variable, representing the number of capacitor groups connected at time t; is the upper limit of the number of group-switched capacitors in operation at node i; in formula (30), The maximum number of switching times of grouped switching capacitors; represents the change in compensation capacity of capacitors switched in groups at adjacent moments; in formula (31), B SVC is a set of nodes including static VAR compensators; and They are the upper and lower limits of the static VAR compensator compensation power respectively; is the compensation capacity of the static VAR compensator at time t.
[0048] Furthermore, the model division method in step S4 is as follows:
[0049]
[0050]
[0051]
[0052]
[0053]
[0054]
[0055] In formula (32), k is the current number of iterations; y l The optimal solution of continuous variables obtained by solving the subproblem after the lth iteration; is the value of the uncertain variable of the subproblem in the worst scenario after the lth iteration; in formula (34), ρ is the optimal solution of the subproblem; α, β, γ, ν, κ, are the Lagrange multipliers in the corresponding dual process respectively; Formula (37) indicates that the main problem is divided into the restricted main problem and the sub-problem, which can be solved by the column and constraint generation algorithm.
[0056] Furthermore, the expression of the intraday optimization dispatching model of the active distribution network in step S5 is as follows:
[0057]
[0058]
[0059]
[0060]
[0061]
[0062]
[0063]
[0064]
[0065]
[0066]
[0067] In formula (38), Optimize the operating cost of active distribution network in the future; B N is a node set; B sub C is the node set connected to the upper grid transformer; t G is the electricity purchase cost in period t; C t ESS is the operating cost of energy storage during period t; C t grid is the transaction price between the active distribution network and the microgrid at time t. When it is positive, it represents the revenue from the active distribution network selling electricity to the microgrid, and when it is negative, it represents the cost of the active distribution network purchasing electricity from the microgrid; λ t G P is the electricity selling price of the upper distribution network during period t; t sub B is the actual active power demand of the distribution network transformer at node i to the upper grid during period t; ESS is the node set where the energy storage is located; c ESS The charging, discharging and maintenance costs of energy storage; and is the charging and discharging power of the energy storage at node i at time t; in formula (39), P ij,t , Q ij,t are the active and reactive power of line ij at time t; v i,t is the node voltage; P i,t , They are respectively the active power input from the upper power grid, the active power output of the distributed generation, the load reduction power, and the load active power at time t; Q i,t , are the reactive power input from the upper grid at time t, the injected power of node i, the reactive power output of distributed generation, the compensation capacity of group switching capacitors, the compensation capacity of static VAR compensator and the reactive power of load; tanφ is the power factor angle tan value of the reduced load. In formula (40), S ij,max is the maximum apparent power of the line; in formula (41), θ ij,t is the phase angle difference between the two ends of the line; g ij 、b ij are the conductance and susceptance of line ij respectively; U i,t is the voltage square term of node i; in formula (42), and are the actual daily output and the day-ahead predicted output of the microgrid connected to node i at time t, respectively; are the actual reactive power of the microgrid connected to node i at time t during the day; in formula (43), and are the energy storage charging and discharging states at time t, respectively, and are 0-1 variables; in formula (44), are the charging and discharging power of energy storage i at time t respectively; and is the upper limit of the power value; P char and P dischar is the lower boundary of the power value; in formula (45), is the rated capacity of energy storage i; Δt is the scheduling time interval; and SOC The upper and lower limits of the state of charge of the energy storage; SOC i,t is the state of charge value of energy storage i at time t; SOC i,0 , SOC i,end , SOC i,set are the initial state of charge value of the energy storage in the dispatch period, the state of charge value at the end of the dispatch period and the set initial state of charge value respectively; in formula (46), and They are the upper and lower limits of the static VAR compensator compensation power respectively; is the compensation capacity of the static VAR compensator at time t.
[0068] Furthermore, the steps of intraday rolling optimization of the active distribution network in step S6 are as follows:
[0069] The scheduling period is reduced from 24 hours before the day to the ΔT period, and the rolling process is as follows:
[0070] A1: Set the initial value, rolling window M = 0, system running time t = t 0 ;
[0071] A2: Read the 2h forecast data at time t-δt, select the scheduling time [t, t+ΔT d ] The model is solved with the goal of minimizing the operating cost, and only the first cycle [t, t+ΔT is executed 1 ]Time schedule;
[0072] A3: Update the initial value, set the rolling window M = M + 1, and the system running time t = t 0 +MΔT r , return to step A2.
[0073] Compared with the prior art, the present invention has the following beneficial effects: the energy management strategy of renewable energy generation accessing the distribution network in the form of microgrid is studied, the uncertainty of renewable energy generation power and the self-management capability of microgrid are considered, and based on the hierarchical scheduling research method, from the time scale of day-ahead-day and microgrid-distribution network, the optimization scheduling problem of active distribution network under the access of renewable energy generation in an uncertain environment is solved. The economic losses caused by uncertainty factors to the system are effectively reduced, a more economical and feasible scheduling plan is provided to the dispatchers, and the self-balancing operation capability of the distribution network is effectively improved. BRIEF DESCRIPTION OF THE DRAWINGS
[0074] Figure 1 It is a flowchart of an uncertainty hierarchical scheduling method for an active distribution network containing multiple microgrids according to the present invention;
[0075] Figure 2 It is a diagram of the power grid structure used in the example of the present invention. DETAILED DESCRIPTION
[0076] The present invention is further explained below in conjunction with the accompanying drawings and specific embodiments. It should be understood that these embodiments are only used to illustrate the present invention and are not used to limit the scope of the present invention. After reading the present invention, various equivalent forms of modifications to the present invention by those skilled in the art all fall within the scope defined by the claims attached to this application.
[0077] like Figure 1 As shown, the present invention provides an uncertainty hierarchical scheduling method for an active distribution network containing multiple microgrids, comprising the following steps:
[0078] S1: Establish a day-ahead optimal dispatch model for microgrids, use the opportunity constraint programming model to describe the forecast error of renewable energy power generation output, and limit the uncertainty interval of intraday interactive power to a certain range with a certain probability level based on economic efficiency and uncertainty absorption capacity;
[0079] The expression of the day-ahead optimal dispatch model of the microgrid is as follows:
[0080]
[0081]
[0082]
[0083]
[0084]
[0085]
[0086]
[0087]
[0088]
[0089]
[0090]
[0091] In formula (1), is the total cost of the microgrid operation on the day before; T is the total dispatching period on the day before; C t grid is the electricity purchase and sale cost between the microgrid and the upper-level distribution network during period t; C t DG is the operating cost of the distributed generation during period t; C t ESS is the operating cost of energy storage during period t; C t f is the fuel consumption cost of the gas turbine during period t; B N is a node set; and are the purchase price and sales price of electricity from the microgrid to the active distribution network in period t, respectively; is the power transmission plan of the active distribution network to the microgrid in period t; is the transmission power plan from the microgrid to the active distribution network at time t; c DG P is the operating cost of distributed power generation; t DG B is the output of distributed power generation during period t; ESS is the node set where the energy storage is located; c ESS The charging, discharging and maintenance costs of energy storage; and is the charging and discharging power of the energy storage at node i at time t. Fis the node set where the gas turbine is located; c f is the fuel consumption cost per unit power of the gas turbine; is the output of the gas turbine at node i at time t; in formula (2), P t grid is the actual value of the interaction power between the active distribution network and the microgrid at time t; is the active power demand of node i during period t; in formula (3), and are the state variables of power transmission from the active distribution network to the microgrid and from the microgrid to the active distribution network at time t, respectively, which are 0-1 variables; in formula (4), Provides an upper limit on the power transmitted from the active distribution network to the microgrid; is the upper limit of the power transmitted from the microgrid to the active distribution network; in formula (5), δ is the preset interactive power prediction error range; in formula (6), The output of distributed power generation is predicted at time t; t is the output prediction error of the distributed generation at time t; in formula (7), σ t is the standard deviation of the normal distribution of the distributed generation prediction error; ρ DG and is the correlation coefficient of the prediction error; P DG is the installed capacity of distributed generation; in formula (8), and P i f are the upper and lower limits of the gas turbine output at node i respectively; in formula (9), u t char and are the energy storage charging and discharging states at time t, respectively, and are 0-1 variables; in formula (10), are the charging and discharging power of energy storage i at time t respectively; and is the upper limit of the power value; P char and P dischar is the lower boundary of the power value; in formula (11), is the rated capacity of energy storage i; Δt is the scheduling time interval; and SOC The upper and lower limits of the state of charge of the energy storage; SOC i,t is the state of charge value of energy storage i at time t; SOC i,0 , SOC i,end , SOC i,set They are the initial state of charge value of the energy storage in the dispatching period, the state of charge value at the end of the dispatching period and the set initial state of charge value.
[0092] S2: Convert and solve the microgrid's day-ahead optimal dispatch model, and report the dispatch plan to the distribution network;
[0093] The expression of the microgrid day-ahead optimization model transformation is as follows:
[0094]
[0095]
[0096]
[0097]
[0098] In formula (12), Φ -1 is the inverse function of Φ; sup is the maximum value of the sequence; in formula (13), K α To satisfy A real number.
[0099] S3: Establish a day-ahead optimal dispatching model for the active distribution network, determine the action state of the discrete dispatching unit, and perform robust optimal dispatching on the distribution network side based on the interactive power uncertainty interval and load forecast information reported by the microgrid;
[0100] The expression of the day-ahead optimal dispatching model of active distribution network is as follows:
[0101]
[0102]
[0103]
[0104]
[0105]
[0106]
[0107]
[0108]
[0109]
[0110]
[0111]
[0112]
[0113]
[0114]
[0115]
[0116]
[0117] In formula (16), Optimize the operating cost of active distribution network in the future; B N is a node set; B sub C is the node set connected to the upper grid transformer; t G is the electricity purchase cost in period t; C t LC is the dispatching cost of load reduction in period t; C t ESS is the operating cost of energy storage during period t; C t grid is the transaction price between the active distribution network and the microgrid at time t. When it is positive, it represents the revenue from the active distribution network selling electricity to the microgrid, and when it is negative, it represents the cost of the active distribution network purchasing electricity from the microgrid; λ t G P is the electricity selling price of the upper distribution network during period t; t sub B is the actual active power demand of the distribution network transformer at node i to the upper grid during period t; LC is the set of nodes where the load can be reduced; c LC Reimbursement costs for demand-side response; is the demand side response load; B ESS is the node set where the energy storage is located; c ESS The charging, discharging and maintenance costs of energy storage; and B is the charging and discharging power of energy storage at node i at time t; MG It is the set of nodes connected to the microgrid; and are the purchase price and sales price of the active distribution network for the microgrid in period t; P t gridin is the transmission power from the active distribution network to the microgrid at node i at time t; P t gridout is the transmission power from the microgrid at node i to the active distribution network at time t; in formula (17), π(j) is the set of branch headend nodes with node j as the terminal node; P j,t and Q j,t are the active injection power and reactive injection power of node j at time t respectively; P ij,t and Q ij,t are respectively the active power and reactive power of branch ij at time t; V i,t is the node voltage of node i at time t; and are respectively the active output and reactive output of the distributed generation connected to node i at time t; and are the active load and reactive load of node i at time t respectively; is the load reduction size of node i at time t; is the reactive power output of the group switching capacitor connected to node i at time t; is the reactive power output of the static VAR compensator connected to node i at time t; tanφ is the tangent value of the power factor angle of the reduced load; in formula (18), is the square of the current of line ij at time t; is the square of the voltage of node i at time t; in formula (19), and P sub are the upper and lower limits of the upper power grid transmission power respectively; in formula (20), and are the actual daily output and the day-ahead predicted output of the microgrid connected to node i at time t; i,t ≥0, is the fluctuation rate of the distributed power supply connected to node i at time t, that is, the maximum fluctuation deviation percentage; δ i is the error range reported by the microgrid connected to node i on the previous day; τ i is the uncertain budget for the output time constraint of distributed generation, which can be used to adjust the conservatism of the solution; in formula (22), and are the energy storage charging and discharging states at time t, respectively, and are 0-1 variables; in formula (23), are the charging and discharging power of energy storage i at time t respectively; and is the upper limit of the power value; P char and P dischar is the lower boundary of the power value; in formula (24), is the rated capacity of energy storage i; Δt is the scheduling time interval; and SOC The upper and lower limits of the state of charge of the energy storage; SOC i,t is the state of charge value of energy storage i at time t; SOC i,0 , SOC i,end , SOC i,set are the initial state of charge value of the energy storage in the dispatching period, the state of charge value at the end of the dispatching period and the set initial state of charge value respectively; in formula (25), is the actual load reduction at time t; is the maximum load reduction that can be achieved at time t; is a 0-1 variable, indicating the load reduction state; in formula (26), is a 0-1 variable, indicating the start of load shedding and the stop of load shedding respectively; in formula (27), They represent the shortest load reduction time and the longest load reduction time respectively; in formula (28), represents the number of loads that can be reduced simultaneously; in formula (29), B CB is a node set including group switching capacitors; is the operational capacity of the capacitor at node i at time t, The gear switching capacity for the gear change of the group switching capacitor; is an integer variable, representing the number of capacitor groups connected at time t; is the upper limit of the number of group-switched capacitors in operation at node i; in formula (30), The maximum number of switching times of grouped switching capacitors; represents the change in compensation capacity of capacitors switched in groups at adjacent moments; in formula (31), B SVC is a set of nodes including static VAR compensators; and They are the upper and lower limits of the static VAR compensator compensation power respectively; is the compensation capacity of the static VAR compensator at time t.
[0118] S4: The column and constraint generation algorithm is used to divide the robust optimization model into a main problem and subproblems with reduced variable scale;
[0119] The model partitioning method is as follows:
[0120]
[0121]
[0122]
[0123]
[0124]
[0125]
[0126] In formula (32), k is the current number of iterations; y l The optimal solution of continuous variables obtained by solving the subproblem after the lth iteration; is the value of the uncertain variable of the subproblem in the worst scenario after the lth iteration; in formula (34), ρ is the optimal solution of the subproblem; α, β, γ, ν, κ, are the Lagrange multipliers in the corresponding dual process respectively; Formula (37) indicates that the main problem is divided into the restricted main problem and the sub-problem, which can be solved by the column and constraint generation algorithm.
[0127] S5: Establish an intraday optimal dispatch model for active distribution networks to improve system behavior;
[0128] The expression of the intraday optimal dispatching model of active distribution network is as follows:
[0129]
[0130]
[0131]
[0132]
[0133]
[0134]
[0135]
[0136]
[0137]
[0138]
[0139] In formula (38), Optimize the operating cost of active distribution network in the future; B N is a node set; B sub C is the node set connected to the upper grid transformer; t G is the electricity purchase cost in period t; C t ESS is the operating cost of energy storage during period t; C t grid is the transaction price between the active distribution network and the microgrid at time t. When it is positive, it represents the revenue from the active distribution network selling electricity to the microgrid, and when it is negative, it represents the cost of the active distribution network purchasing electricity from the microgrid; λ t G P is the electricity selling price of the upper distribution network during period t; t sub B is the actual active power demand of the distribution network transformer at node i to the upper grid during period t; ESS is the node set where the energy storage is located; c ESS The charging, discharging and maintenance costs of energy storage; and is the charging and discharging power of the energy storage at node i at time t; in formula (39), P ij,t , Q ij,t are the active and reactive power of line ij at time t; v i,t is the node voltage; P i,t , They are respectively the active power input from the upper power grid, the active power output of the distributed generation, the load reduction power, and the load active power at time t; Q i,t , are the reactive power input from the upper grid at time t, the injected power of node i, the reactive power output of distributed generation, the compensation capacity of group switching capacitors, the compensation capacity of static VAR compensator and the reactive power of load; tanφ is the power factor angle tan value of the reduced load. In formula (40), S ij,max is the maximum apparent power of the line; in formula (41), θ ij,t is the phase angle difference between the two ends of the line; g ij 、b ij are the conductance and susceptance of line ij respectively; U i,t is the voltage square term of node i; in formula (42), and are the actual daily output and the day-ahead predicted output of the microgrid connected to node i at time t, respectively; are the actual reactive power of the microgrid connected to node i at time t during the day; in formula (43), and are the energy storage charging and discharging states at time t, respectively, and are 0-1 variables; in formula (44), are the charging and discharging power of energy storage i at time t respectively; and is the upper limit of the power value; P char and P dischar is the lower boundary of the power value; in formula (45), is the rated capacity of energy storage i; Δt is the scheduling time interval; and SOC The upper and lower limits of the state of charge of the energy storage; SOC i,t is the state of charge value of energy storage i at time t; SOC i,0 , SOC i,end , SOC i,set are the initial state of charge value of the energy storage in the dispatch period, the state of charge value at the end of the dispatch period and the set initial state of charge value respectively; in formula (46), and They are the upper and lower limits of the static VAR compensator compensation power respectively; is the compensation capacity of the static VAR compensator at time t.
[0140] S6: Use intraday rolling optimization to determine the action volume of the continuous scheduling unit within the day.
[0141] The steps for active distribution network intraday rolling optimization are as follows:
[0142] The scheduling period is reduced from 24 hours before the day to the ΔT period, and the rolling process is as follows:
[0143] A1: Set the initial value, rolling window M = 0, system running time t = t 0 ;
[0144] A2: Read the 2h forecast data at time t-δt, select the scheduling time [t, t+ΔT d ] The model is solved with the goal of minimizing the operating cost, and only the first cycle [t, t+ΔT is executed 1 ]Time schedule;
[0145] A3: Update the initial value, set the rolling window M = M + 1, and the system running time t = t 0 +MΔT r , return to step A2.
[0146] Based on the above content, in order to verify the effectiveness of the solution of the present invention, the above solution is applied as an example in this embodiment, as follows:
[0147] A 51-node 10kV distribution network test system in a rural area was selected for improvement. Figure 2 The node load and feeder parameters are shown in Table 1 and Table 2.
[0148] Table 1 Test system node load size
[0149]
[0150] Table 2 Test system feeder parameters
[0151]
[0152]
[0153] In order to verify the uncertainty hierarchical management framework proposed in the present invention, the confidence level in the microgrid day-ahead dispatch model is set to 95%, the uncertainty budget of the distribution network robust optimization model is set to 6, and the following scenarios are set for cost comparison analysis.
[0154] Scenario 1: The uncertainty of PV output is entirely borne by the microgrid;
[0155] Scenario 2: The microgrid limits the fluctuation range of the tie line power to within 5%, and the tie line fluctuation within 5% is borne by the distribution network;
[0156] Scenario 3: The microgrid limits the fluctuation range of the interconnection line power to within 10%, and the interconnection line fluctuation within 10% is borne by the distribution network.
[0157] The results in several scenarios are shown in Table 3.
[0158] Table 3 Scheduling results under different scenarios
[0159]
[0160]
[0161] By analyzing the results of the example, it is not difficult to find that as the uncertainty range of the interactive power reported by the microgrid continues to increase, its own operating cost increases significantly. Specifically, the uncertainty of renewable energy generation leads to fluctuations in the power demand of the microgrid and the power sales of the microgrid, which increases the demand for power purchases from the upper power grid, and significantly reduces the economic benefits. At the same time, by comparing the network loss indicators under several scenarios, it can be found that as the power purchase power of the distribution network from the microgrid increases, the network loss in the distribution network line continues to increase. Analyzing the reasons, it can be found that the power supply of the microgrid to the distribution network causes a reverse flow, resulting in an increase in the system network loss. The results of this example show that the fluctuation of the tie line power is proportional to the operating cost of the distribution network, proving the necessity of uncertainty hierarchical dispatch management measures when the microgrid participates in the power market as an independent entity.
[0162] Table 4 compares the distribution network operation costs of two schemes: inheriting all operation plans and only inheriting discrete variable plans during the intraday rolling stage. The results show the effectiveness of the intraday rolling optimization scheme proposed in this paper in achieving economic operation of the distribution network.
[0163] Table 4 Operational indicators under different succession plans
[0164]
[0165] According to the results of this embodiment, on the basis of the hierarchical scheduling framework, the uncertainty of the output of distributed renewable power sources is first described by the opportunity constraint planning condition, and then the uncertainty of the interactive power plan reported by the microgrid is described by the polyhedron uncertainty set. Finally, the output plan of the continuous scheduling equipment is adjusted within the day through the rolling optimization theory, which promotes the penetration of distributed power sources and greatly reduces the uncertainty factors in the distribution network. In summary, the beneficial effect of the research on the uncertainty hierarchical scheduling of the active distribution network containing multiple microgrids provided by the present invention is reflected.
[0166] Finally, it should be noted that the above embodiments are only used to illustrate the technical solution of the present invention and are not intended to limit the present invention. A person skilled in the art should understand that the specific implementation of the present invention can be modified or replaced by equivalents, but these modifications or changes are within the scope of protection of the pending claims.
Claims
1. An uncertainty hierarchical scheduling method for an active distribution network containing multiple microgrids, characterized in that: The method comprises the following steps: S1: Establish a day-ahead optimal dispatch model for microgrids, use the opportunity constraint programming model to describe the forecast error of renewable energy power generation output, and limit the uncertainty interval of intraday interactive power to a certain range with a certain probability level based on economic efficiency and uncertainty absorption capacity; S2: Convert and solve the microgrid's day-ahead optimal dispatch model, and report the dispatch plan to the distribution network; S3: Establish a day-ahead optimal dispatching model for the active distribution network, determine the action state of the discrete dispatching unit, and perform robust optimal dispatching on the distribution network side based on the interactive power uncertainty interval and load forecast information reported by the microgrid; S4: The column and constraint generation algorithm is used to divide the robust optimization model into a main problem and subproblems with reduced variable scale; S5: Establish an intraday optimal dispatch model for active distribution networks to improve system behavior; S6: Determine the action amount of the continuous scheduling unit within the day by using intraday rolling optimization; The expression of the day-ahead optimization dispatch model of the microgrid in step S1 is as follows: In formula (1), is the total cost of the microgrid operation on the day before; T is the total dispatching period on the day before; C t grid is the electricity purchase and sale cost between the microgrid and the upper-level distribution network during period t; C t DG is the operating cost of the distributed generation during period t; C t ESS is the operating cost of energy storage during period t; C t f is the fuel consumption cost of the gas turbine during period t; B N is a node set; and are the purchase price and sales price of electricity from the microgrid to the active distribution network in period t, respectively; is the power transmission plan of the active distribution network to the microgrid in period t; is the transmission power plan from the microgrid to the active distribution network at time t; c DG P is the operating cost of distributed power generation; t DG B is the output of distributed power generation during period t; ESS is the node set where the energy storage is located; c ESS The charging, discharging and maintenance costs of energy storage; and B is the charging and discharging power of energy storage at node i at time t; F is the node set where the gas turbine is located; c f is the fuel consumption cost per unit power of the gas turbine; is the output of the gas turbine at node i at time t; in formula (2), P t grid is the actual value of the interaction power between the active distribution network and the microgrid at time t; is the active power demand of node i during period t; in formula (3), and are the state variables of power transmission from the active distribution network to the microgrid and from the microgrid to the active distribution network at time t, respectively, which are 0-1 variables; in formula (4), Provides an upper limit on the power transmitted from the active distribution network to the microgrid; is the upper limit of the power transmitted from the microgrid to the active distribution network; in formula (5), δ is the preset interactive power prediction error range; in formula (6), The output of distributed power generation is predicted at time t; t is the output prediction error of the distributed generation at time t; in formula (7), σ t is the standard deviation of the normal distribution of the distributed generation prediction error; ρ DG and is the correlation coefficient of the prediction error; P DG is the installed capacity of distributed generation; in formula (8), and P i f are the upper and lower limits of the gas turbine output at node i respectively; in formula (9), and are the energy storage charging and discharging states at time t, respectively, and are 0-1 variables; in formula (10), are the charging and discharging power of energy storage i at time t respectively; and is the upper limit of the power value; P char and P dischar is the lower boundary of the power value; in formula (11), is the rated capacity of energy storage i; Δt is the scheduling time interval; and SOC The upper and lower limits of the state of charge of the energy storage; SOC i,t is the state of charge value of energy storage i at time t; SOC i,0 , SOC i,end , SOC i,set They are the initial state of charge value of the energy storage in the dispatch period, the state of charge value at the end of the dispatch period and the set initial state of charge value; The expression for the transformation of the microgrid day-ahead optimization model in step S2 is as follows: In formula (12), Φ -1 is the inverse function of Φ; sup is the maximum value of the sequence; in formula (13), K α To satisfy a real number; The expression of the active distribution network day-ahead optimization dispatching model in step S3 is as follows: In formula (16), Optimize the operating cost of active distribution network in the future; B N is a node set; B sub C is the node set connected to the upper grid transformer; t G is the electricity purchase cost in period t; C t LC is the dispatching cost of load reduction in period t; C t ESS is the operating cost of energy storage during period t; C t grid is the transaction price between the active distribution network and the microgrid at time t. When it is positive, it represents the revenue from the active distribution network selling electricity to the microgrid, and when it is negative, it represents the cost of the active distribution network purchasing electricity from the microgrid; λ t G P is the electricity selling price of the upper distribution network during period t; t sub B is the actual active power demand of the distribution network transformer at node i to the upper grid during period t; LC is the set of nodes where the load can be reduced; c LC Reimbursement costs for demand-side response; is the demand side response load; B ESS is the node set where the energy storage is located; c ESS The charging, discharging and maintenance costs of energy storage; and B is the charging and discharging power of energy storage at node i at time t; MG It is the set of nodes connected to the microgrid; and are the purchase price and sales price of the active distribution network for the microgrid in period t; P t gridin is the transmission power from the active distribution network to the microgrid at node i at time t; P t gridout is the transmission power from the microgrid at node i to the active distribution network at time t; in formula (17), π(j) is the set of branch headend nodes with node j as the terminal node; P j,t and Q j,t are the active injection power and reactive injection power of node j at time t respectively; P ij,t and Q ij,t are respectively the active power and reactive power of branch ij at time t; V i,t is the node voltage of node i at time t; and are respectively the active output and reactive output of the distributed generation connected to node i at time t; and are the active load and reactive load of node i at time t respectively; is the load reduction size of node i at time t; is the reactive power output of the group switching capacitor connected to node i at time t; is the reactive power output of the static VAR compensator connected to node i at time t; tanφ is the tangent value of the power factor angle of the reduced load; in formula (18), is the square of the current of line ij at time t; is the square of the voltage of node i at time t; in formula (19), and P sub are the upper and lower limits of the upper power grid transmission power respectively; in formula (20), and are the actual daily output and the day-ahead predicted output of the microgrid connected to node i at time t; i,t ≥0, is the fluctuation rate of the distributed power supply connected to node i at time t, that is, the maximum fluctuation deviation percentage; δ i is the error range reported by the microgrid connected to node i on the previous day; τ i is the uncertain budget for the output time constraint of distributed generation, which can be used to adjust the conservatism of the solution; in formula (22), and are the energy storage charging and discharging states at time t, respectively, and are 0-1 variables; in formula (23), are the charging and discharging power of energy storage i at time t respectively; and is the upper limit of the power value; P char and P dischar is the lower boundary of the power value; in formula (24), is the rated capacity of energy storage i; Δt is the scheduling time interval; and SOC The upper and lower limits of the state of charge of the energy storage; SOC i,t is the state of charge value of energy storage i at time t; SOC i,0 , SOC i,end , SOC i,set are the initial state of charge value of the energy storage in the dispatching period, the state of charge value at the end of the dispatching period and the set initial state of charge value respectively; in formula (25), is the actual load reduction at time t; is the maximum load reduction that can be achieved at time t; is a 0-1 variable, indicating the load reduction state; in formula (26), is a 0-1 variable, indicating the start of load shedding and the stop of load shedding respectively; in formula (27), They represent the shortest load reduction time and the longest load reduction time respectively; in formula (28), represents the number of loads that can be reduced simultaneously; in formula (29), B CB is a node set including group switching capacitors; is the operational capacity of the capacitor at node i at time t, The gear switching capacity for the gear change of the group switching capacitor; is an integer variable, representing the number of capacitor groups connected at time t; is the upper limit of the number of group-switched capacitors in operation at node i; in formula (30), The maximum number of switching times of grouped switching capacitors; represents the change in compensation capacity of capacitors switched in groups at adjacent moments; in formula (31), B SVC is a set of nodes including static VAR compensators; and They are the upper and lower limits of the static VAR compensator compensation power respectively; is the compensation capacity of the static VAR compensator at time t.
2. The uncertainty hierarchical scheduling method for active distribution network containing multiple microgrids according to claim 1 is characterized in that: The model division method in step S4 is as follows: In formula (32), k is the current number of iterations; y l The optimal solution of continuous variables obtained by solving the subproblem after the lth iteration; is the value of the uncertain variable of the subproblem in the worst scenario after the lth iteration; in formula (34), ρ is the optimal solution of the subproblem; α, β, γ, ν, κ, θ are the Lagrange multipliers in the corresponding dual process respectively; formula (37) indicates that the main problem is divided into the restricted main problem and the subproblem, which can be solved by the column and constraint generation algorithm.
3. The uncertainty hierarchical scheduling method for active distribution network containing multiple microgrids according to claim 1 is characterized in that: The expression of the intraday optimization dispatching model of the active distribution network in step S5 is as follows: In formula (38), Optimize the operating cost of active distribution network in the future; B N is a node set; B sub C is the node set connected to the upper grid transformer; t G is the electricity purchase cost in period t; C t ESS is the operating cost of energy storage during period t; C t grid is the transaction price between the active distribution network and the microgrid at time t. When it is positive, it represents the revenue from the active distribution network selling electricity to the microgrid, and when it is negative, it represents the cost of the active distribution network purchasing electricity from the microgrid; λ t G P is the electricity selling price of the upper distribution network during period t; t sub B is the actual active power demand of the distribution network transformer at node i to the upper grid during period t; ESS is the node set where the energy storage is located; c ESS The charging, discharging and maintenance costs of energy storage; and is the charging and discharging power of the energy storage at node i at time t; in formula (39), P ij,t , Q ij,t are the active and reactive power of line ij at time t; v i,t is the node voltage; P i,t , They are respectively the active power input from the upper grid at time t, the active power output of the distributed generation, the load reduction power, and the load active power; Q t sub , Q i,t , are the reactive power input from the upper grid at time t, the injected power of node i, the reactive power output of distributed generation, the compensation capacity of group switching capacitors, the compensation capacity of static VAR compensator and the reactive power of load; tanφ is the power factor angle tan value of the reduced load; In formula (40), S ij,max is the maximum apparent power of the line; in formula (41), θ ij,t is the phase angle difference between the two ends of the line; g ij 、b ij are the conductance and susceptance of line ij respectively; U i,t is the voltage square term of node i; in formula (42), and are the actual daily output and the day-ahead predicted output of the microgrid connected to node i at time t, respectively; are the actual reactive power of the microgrid connected to node i at time t during the day; in formula (43), and are the energy storage charging and discharging states at time t, respectively, and are 0-1 variables; in formula (44), are the charging and discharging power of energy storage i at time t respectively; and is the upper limit of the power value; P char and P dischar is the lower limit of the power value; in formula (45), E i nomal is the rated capacity of energy storage i; Δt is the scheduling time interval; and SOC The upper and lower limits of the state of charge of the energy storage; SOC i,t is the state of charge value of energy storage i at time t; SOC i,0 , SOC i,end , SOC i,set are the initial state of charge value of the energy storage in the dispatch period, the state of charge value at the end of the dispatch period and the set initial state of charge value respectively; in formula (46), and They are the upper and lower limits of the static VAR compensator compensation power respectively; is the compensation capacity of the static VAR compensator at time t.
4. The method for uncertain hierarchical scheduling of active distribution network containing multiple microgrids according to claim 1 is characterized in that: The steps of intraday rolling optimization of the active distribution network in step S6 are as follows: The scheduling period is reduced from 24 hours before the day to the ΔT period, and the rolling process is as follows: A1: Set the initial value, rolling window M = 0, system running time t = t0; A2: Read the 2h forecast data at time t-δt, select the scheduling time [t, t+ΔT d ] The model is solved with the goal of minimizing the operating cost, and only the scheduling plan for the first cycle, i.e., [t, t+ΔT1], is executed; A3: Update the initial value, set the rolling window M = M + 1, and the system running time t = t0 + MΔT r , return to step A2.
