Distribution network dispatching method and system considering the uncertainty of wind, solar and electric vehicles

Abstract

The present invention provides a distribution network dispatching method and system considering the uncertainty of wind, solar and electric vehicles, belonging to the field of distribution network optimization dispatching, and the distribution network dispatching method includes: determining the fast charging station cluster constraint according to the charging and discharging conditions of each electric vehicle in the fast charging station cluster at a historical moment; determining the first-stage objective function and operation constraint conditions according to the day-ahead forecast data of conventional units, wind turbines, and photovoltaic power stations and the load forecast value of each node; based on the day-ahead forecast data of wind turbines and photovoltaic power stations, using a generative adversarial network to generate a real-time scenario set; determining the second-stage objective function and operation constraint conditions according to the real-time scenario set; establishing and solving a two-stage random optimization dispatching model according to the two-stage objective functions, fast charging station cluster constraints and two-stage operation constraints, and obtaining the dispatching result of the distribution network. It promotes the consumption of renewable energy and improves the power balance of the distribution network.

Description

Technical Field

[0001] The present invention relates to the field of distribution network optimization and dispatching, and in particular to a distribution network dispatching method and system taking into account the uncertainties of wind, solar and electric vehicles. Background Art

[0002] With the adjustment of the global energy structure and the rapid development of clean energy such as wind power and solar energy, the global production and sales of electric vehicles have grown rapidly. The development of electrified transportation provides a low-carbon, environmentally friendly, economical and efficient solution to future transportation power problems, and can also alleviate the fossil energy supply crisis and environmental pollution problems in the transportation system. The electrified transportation system is an important load of the power system, and the development of electrified transportation has made the interaction between the power system and the transportation system increasingly significant.

[0003] The large-scale development of electric vehicles will lead to the innovation of transportation electrification. Due to the large amount of uncertainty in user behavior and the uncertainty in the output of new energy such as wind and solar power, it brings challenges to the power balance and scheduling of the power system. Summary of the invention

[0004] The purpose of the present invention is to provide a distribution network scheduling method and system taking into account the uncertainties of wind, solar and electric vehicles, which can improve the power balance of the distribution network and promote the consumption of renewable energy.

[0005] To achieve the above object, the present invention provides the following solutions:

[0006] A distribution network dispatching method considering the uncertainty of wind, solar and electric vehicles, including:

[0007] According to the charging and discharging conditions of each electric vehicle in the fast charging station cluster at a historical moment, the fast charging station cluster constraints are determined;

[0008] Obtain the day-ahead forecast data of conventional units in the distribution network, the day-ahead forecast data of wind turbines, the day-ahead forecast data of photovoltaic power stations, and the load forecast value of each node;

[0009] Determine the first-stage objective function based on the day-ahead forecast data of conventional units;

[0010] Determine the operating constraints of the first phase based on the day-ahead forecast data of conventional units, the day-ahead forecast data of wind turbines, the day-ahead forecast data of photovoltaic power stations, and the load forecast values ​​of each node;

[0011] Based on the day-ahead forecast data of the wind turbine and the day-ahead forecast data of the photovoltaic power station, a generative adversarial network is used to generate a real-time scenario set; the real-time scenario set includes multiple real-time scenarios, each of which includes the active output of the wind turbine and the active output of the photovoltaic power station;

[0012] Determining the second-stage objective function and the second-stage operation constraints according to the real-time scenario set;

[0013] Establishing a two-stage stochastic optimization scheduling model according to the first-stage objective function, the second-stage objective function, the fast charging station cluster constraint, the first-stage operation constraint condition and the second-stage operation constraint condition;

[0014] The two-stage random optimization scheduling model is solved to obtain the scheduling results of the distribution network; the scheduling results include the combination of conventional units, the reserve capacity plan, the output of conventional units, the power adjustment of wind turbines and the power adjustment of photovoltaic power stations.

[0015] To achieve the above object, the present invention also provides the following solution:

[0016] A distribution network dispatching system considering the uncertainty of wind, solar and electric vehicles, comprising:

[0017] A cluster constraint determination unit, used to determine the fast charging station cluster constraint according to the charging and discharging conditions of each electric vehicle in the fast charging station cluster at a historical moment;

[0018] A day-ahead forecast data acquisition unit is used to acquire the day-ahead forecast data of conventional units in the distribution network, the day-ahead forecast data of wind turbines, the day-ahead forecast data of photovoltaic power stations, and the load forecast value of each node;

[0019] A first target determination unit, connected to the day-ahead forecast data acquisition unit, is used to determine a first-stage target function according to the day-ahead forecast data of the conventional unit;

[0020] a first constraint determination unit connected to the day-ahead forecast data acquisition unit, and configured to determine the first-stage operation constraint conditions according to the day-ahead forecast data of the conventional unit, the day-ahead forecast data of the wind turbine unit, the day-ahead forecast data of the photovoltaic power station, and the load forecast value of each node;

[0021] A real-time scenario generation unit is connected to the day-ahead prediction data acquisition unit, and is used to generate a real-time scenario set based on the day-ahead prediction data of the wind turbine generator set and the day-ahead prediction data of the photovoltaic power station by using a generative adversarial network; the real-time scenario set includes a plurality of real-time scenarios, each of which includes the active output of the wind turbine generator set and the active output of the photovoltaic power station;

[0022] A second objective and constraint determination unit, connected to the real-time scenario generation unit, for determining a second-stage objective function and second-stage operation constraints according to the real-time scenario set;

[0023] a model building unit connected to the cluster constraint determination unit, the first target determination unit, the first constraint determination unit, and the second target and constraint determination unit, and configured to establish a two-stage stochastic optimization scheduling model according to the first-stage objective function, the second-stage objective function, the fast charging station cluster constraint, the first-stage operation constraint condition, and the second-stage operation constraint condition;

[0024] A solving unit is connected to the model to solve the two-stage random optimization scheduling model to obtain a scheduling result of the distribution network.

[0025] According to the specific embodiment provided by the present invention, the present invention discloses the following technical effects: first, according to the charging and discharging conditions of each electric vehicle in the fast charging station cluster at a historical moment, the fast charging station cluster constraint is determined, and then according to the day-ahead prediction data of the conventional unit, the first-stage objective function is determined, and according to the day-ahead prediction data of the conventional unit, the day-ahead prediction data of the wind turbine unit, the day-ahead prediction data of the photovoltaic power station and the load prediction value of each node, the first-stage operation constraint condition is determined, and then based on the day-ahead prediction data of the wind turbine unit and the day-ahead prediction data of the photovoltaic power station, a generative adversarial network is used to generate a real-time scenario set, and according to the real-time scenario set, the second-stage objective function and the second-stage operation constraint condition are determined, and the fast charging station cluster constraint, the objective function of the two stages and the operation constraint condition are integrated to establish a random optimization scheduling model, and finally the two-stage random optimization scheduling model is solved to determine the combination mode of the conventional unit, the spare capacity plan, the output of the conventional unit, the power adjustment amount of the wind turbine unit and the power adjustment amount of the photovoltaic power station, taking into account the charging demand of electric vehicles and the uncertainty of wind and solar power generation, promoting the consumption of renewable energy, and improving the power balance of the distribution network. BRIEF DESCRIPTION OF THE DRAWINGS

[0026] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative labor.

[0027] Figure 1 This is a flow chart of a distribution network dispatching method considering the uncertainty of wind, solar and electric vehicles according to the present invention;

[0028] Figure 2 Schematic diagram of the optimal dispatching process for the distribution network;

[0029] Figure 3 This is a module schematic diagram of the distribution network dispatching system of the present invention that takes into account the uncertainties of wind, solar and electric vehicles.

[0030] Explanation of symbols:

[0031] Cluster constraint determination unit-1, day-ahead forecast data acquisition unit-2, first target determination unit-3, first constraint determination unit-4, real-time scenario generation unit-5, second target and constraint determination unit-6, model building unit-7, solution unit-8. DETAILED DESCRIPTION

[0032] 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.

[0033] The purpose of the present invention is to provide a distribution network scheduling method and system that takes into account the uncertainties of wind, solar and electric vehicles, so as to promote the consumption of renewable energy and improve the power balance of the distribution network by considering the charging demand of electric vehicles and the uncertainty of wind and solar power generation.

[0034] In order to make the above-mentioned objects, features and advantages of the present invention more obvious and easy to understand, the present invention is further described in detail below with reference to the accompanying drawings and specific embodiments.

[0035] like Figure 1 and Figure 2 As shown, the distribution network dispatching method considering the uncertainty of wind, solar and electric vehicles of the present invention includes:

[0036] S1: Determine the fast charging station cluster constraints based on the charging and discharging conditions of each electric vehicle in the fast charging station cluster at a historical moment.

[0037] S2: Obtain the day-ahead forecast data of conventional units in the distribution network, the day-ahead forecast data of wind turbines, the day-ahead forecast data of photovoltaic power stations, and the load forecast value of each node.

[0038] S3: Determine the first-stage objective function based on the day-ahead forecast data of conventional units.

[0039] S4: Determine the operating constraints of the first stage according to the day-ahead forecast data of conventional units, the day-ahead forecast data of wind turbine units, the day-ahead forecast data of photovoltaic power stations and the load forecast value of each node.

[0040] S5: Based on the day-ahead prediction data of the wind turbine and the day-ahead prediction data of the photovoltaic power station, a generative adversarial network is used to generate a real-time scenario set; the real-time scenario set includes multiple real-time scenarios, each real-time scenario includes the active output of the wind turbine and the active output of the photovoltaic power station.

[0041] S6: Determine the second-stage objective function and the second-stage operation constraints according to the real-time scenario set.

[0042] S7: Establish a two-stage stochastic optimization scheduling model according to the first-stage objective function, the second-stage objective function, the fast charging station cluster constraints, the first-stage operating constraints, and the second-stage operating constraints.

[0043] S8: Solve the two-stage random optimization scheduling model to obtain the scheduling results of the distribution network; the scheduling results include the combination of conventional units, reserve capacity plan, output of conventional units, power adjustment of wind turbines and power adjustment of photovoltaic power stations. Specifically, the two-stage random optimization scheduling model considering the uncertainty of wind, solar and electric vehicles is solved by programming using software YALMIO and CPLEX.

[0044] Further, step S1 includes:

[0045] S11: For any electric vehicle in the fast charging station cluster, determine the maximum charge state of the electric vehicle according to the initial power of the electric vehicle when it is connected to the fast charging station cluster, the maximum charge and discharge power of the electric vehicle, and the maximum power that the electric vehicle can reach.

[0046] S12: Determine the minimum state of charge of the electric vehicle based on the initial power of the electric vehicle when it connects to the fast charging station cluster, the minimum charge and discharge power of the electric vehicle, the moment when the electric vehicle leaves the fast charging station cluster, the expected state of charge when the electric vehicle leaves the fast charging station cluster, the maximum charge and discharge power of the electric vehicle, and the minimum power that the electric vehicle can reach.

[0047] For the dispatchable electric vehicle charging demand, the electric vehicle fast charging station cluster model is used to represent it. The electric vehicle fast charging station cluster model is established based on the travel probability distribution data of electric vehicles. It is assumed that electric vehicles use plug-in charging and support V2G (bidirectional power transmission), and there are enough charging facilities in the fast charging station cluster.

[0048] The charging energy constraint of a single electric vehicle is:

[0049]

[0050] in, is the power of the i-th electric car at time t, is the charging power of the i-th electric car at time t, and Δt is the time interval between time t and time t+1.

[0051] The energy upper and lower bounds of a single electric vehicle are:

[0052]

[0053]

[0054] in, is the maximum state of charge of the ith electric vehicle at time t, is the minimum state of charge of the ith electric vehicle at time t, T0 is the starting time when the electric vehicle is connected to the fast charging station cluster, is the initial power of the ith electric car at time T0, P i cmax is the maximum charge and discharge power of the i-th electric vehicle, P i cmin is the minimum charging and discharging power of the i-th electric vehicle, is the minimum power that the i-th electric car can reach at time t, is the maximum power that the i-th electric car can reach at time t, T d When the electric car is fully charged and leaves the fast charging station cluster, is the charge state that the electric vehicle is expected to achieve when the i-th electric vehicle leaves the fast charging station cluster.

[0055] The comprehensive equation for the state of charge of an electric vehicle is:

[0056]

[0057] in, is the upper bound of the power of the ith electric car at time t, is the lower bound of the power of the ith electric car at time t, is the power of the i-th electric car at time t, is the power of the i-th electric car at time t.

[0058] S13: Determine the maximum state of charge of the fast charging station cluster according to the maximum state of charge of each electric vehicle in the fast charging station cluster.

[0059] S14: Determine the minimum state of charge of the fast charging station cluster according to the minimum state of charge of each electric vehicle in the fast charging station cluster.

[0060] S15: Determine the maximum power of the fast charging station cluster according to the upper limit of the power of each electric vehicle in the fast charging station cluster.

[0061] S16: Determine the minimum power of the fast charging station cluster according to the lower bound of the power of each electric vehicle in the fast charging station cluster.

[0062] By sampling the travel probability and other related parameters of the electric vehicle group, the upper and lower bounds of the state of charge and power of each electric vehicle can be obtained using the above formula. The large-scale sampling results are accumulated on the state of charge and power to obtain the key parameter boundaries of the fast charging station cluster model:

[0063]

[0064]

[0065]

[0066]

[0067] in, is the maximum state of charge of fast charging cluster k at time t, is the minimum state of charge of fast charging cluster k at time t, N e is the number of electric vehicles, is the maximum power of fast charging cluster k at time t, is the minimum power of fast charging cluster k at time t.

[0068] S17: Determine a fast charging station cluster constraint according to a maximum state of charge of the fast charging station cluster, a minimum state of charge of the fast charging station cluster, a maximum power of the fast charging station cluster, and a minimum power of the fast charging station cluster.

[0069] Assuming that the charging efficiency of each electric vehicle is the same, the fast charging station cluster constraint is:

[0070]

[0071]

[0072]

[0073] in, is the overall power of the fast charging station cluster during period t; is the overall charge state of the fast charging station cluster during period t.

[0074] Furthermore, in step S3, the day-ahead forecast data of the conventional units include: startup cost, shutdown cost, power generation cost, reserve price increase, reserve price decrease, reserve quantity increase, and reserve quantity decrease.

[0075] The following formula is used to determine the first-stage objective function:

[0076]

[0077] f g (P g,t )=a g (P g,t ) 2 +b g P g,t +c g ;

[0078] Among them, F1 is the objective function value of the first stage, t is the time, g represents the conventional unit, is the startup cost of conventional unit g, is the downtime cost of conventional unit g, u g,t is a binary variable, u g,t The value of 1 indicates that the conventional unit g is in the startup process at time t, u g,t A value of 0 indicates that the conventional unit g is not in the startup process at time t. g,t is a binary variable, v g,t The value of 1 indicates that the conventional unit g is in the shutdown process at time t, v g,t A value of 0 indicates that the conventional unit g is not in the shutdown process at time t. g (P g,t ) is the power generation cost of conventional unit g, P g,t is the output power of conventional unit g at time t, a g , b g 、c g is the power generation cost coefficient of conventional unit g, The reserve price of conventional unit g is increased. The reserve price of conventional unit g is lowered. is the reserve capacity of conventional unit g at time t, It is the reserve capacity of conventional unit g at time t.

[0079] Furthermore, in step S4, the first stage operation constraints include the first system power balance constraint, conventional unit output and spare capacity constraint, conventional unit start and stop constraint, conventional unit start and stop time constraint, conventional unit ramp constraint and network security constraint.

[0080] The day-ahead forecast data of the conventional unit includes the operating status, output upper limit, output lower limit, output power, upward reserve, downward reserve, maximum downward reserve capacity and maximum upward reserve capacity of the conventional unit. The day-ahead forecast data of the wind turbine unit includes the output forecast value of the wind turbine unit. The day-ahead forecast data of the photovoltaic power station includes the output forecast value of the photovoltaic power station.

[0081] The first system power balance constraint is:

[0082]

[0083] Among them, g represents conventional units, w represents wind turbines, p represents photovoltaic power stations, j represents nodes in the distribution network, k represents charging station clusters, and P g,t is the output power of conventional unit g at time t, is the predicted output value of wind turbine w at time t, is the output prediction value of the photovoltaic power station p at time t, L j,t is the load forecast value of node j in the distribution network at time t, is the overall power of charging station cluster k at time t.

[0084] The conventional unit output and reserve capacity constraints are:

[0085]

[0086] Among them, i g,t is the operating status of the conventional unit g at time t, which is a 0-1 variable, i g,t =1 means that the conventional unit g is in operation at time t, i g,t =0 means that the conventional unit g is not running at time t. is the output upper limit of conventional unit g, is the lower limit of the output of the conventional unit g, P g,t is the output power of conventional unit g at time t, is the reserve capacity of conventional unit g at time t, is the reserve capacity of conventional unit g at time t, is the maximum down-regulated reserve capacity of conventional unit g at time t, It is the maximum upward reserve capacity of conventional unit g at time t.

[0087] Conventional unit start and stop constraints are:

[0088] u g,t -v g,t =i g,t -i g,(t-1) ;

[0089] u g,t +v g,t ≤1.

[0090] The start and stop time constraints of conventional units are:

[0091] (H on,g,(t-1) -T on,g )·(i g,(t-1) -i g,t )≥0;

[0092] (Hoff,g,(t-1) -T off,g )·(i g,t -i g,(t-1) )≥0;

[0093] Among them, H on,g,t is the time that the conventional unit g has been continuously powered on at time t, H off,g,t is the continuous shutdown time of the conventional unit g at time t, T on,g is the minimum startup time of conventional unit g, T off,g It is the minimum startup and shutdown time of conventional unit g.

[0094] The climbing constraints of conventional units are:

[0095]

[0096]

[0097] The network security constraints are:

[0098]

[0099] Among them, F l is the transmission power limit of line l in the distribution network, ζ gl is the power distribution factor from conventional unit g to line l, ζ wl is the power distribution factor from wind turbine w to line l, ζ pl is the power distribution factor from PV plant p to line l, ζ jl is the power distribution factor from node j to line l in the distribution network, ζ kl is the power distribution factor from fast charging station cluster k to line l.

[0100] Furthermore, in step S5, based on the day-ahead forecast data and forecast deviation probability model of wind turbines and photovoltaic power stations, a large number of real-time scenarios of wind turbines and photovoltaic power stations output are generated by generative network adversarial method, and the scenarios are reduced. Specifically, it includes:

[0101] S51: Based on the day-ahead forecast data of the wind turbines and the day-ahead forecast data of the photovoltaic power station, a generative adversarial network is used to generate a preliminary real-time scenario set. The preliminary real-time scenario set includes a preliminary wind power generation scenario class and a preliminary photovoltaic power generation scenario class. The preliminary wind power generation scenario class includes multiple wind power generation scenarios, and the preliminary photovoltaic power generation scenario class includes multiple photovoltaic power generation scenarios.

[0102] Specifically, the day-ahead forecast data and forecast deviation probability model of wind turbine and photovoltaic power station output are processed based on the generative adversarial network to obtain wind power generation scenario S and photovoltaic power generation scenario Y. The generative adversarial network includes generator and discriminator models:

[0103] The discriminator objective function is:

[0104]

[0105] Among them, G is the generator, D is the discriminator, E(·) is the mathematical expectation, P data is the real data distribution, P gen To generate data distribution, D(·) is the discriminator output, V(G,D) is P data and P gen The divergence of x is the day-ahead forecast data.

[0106] The optimal generator can be expressed as:

[0107]

[0108] Among them, G * is the optimal generator.

[0109] S52: Use the K-medoids method to reduce the preliminary real-time scene set to obtain the final real-time scene set.

[0110] Specifically, based on the wind power generation scenario S and photovoltaic power generation scenario Y generated by the generative adversarial network, the wind power generation scenario and photovoltaic power generation scenario are reduced respectively. The objective function of the K-medoids scenario reduction process is:

[0111]

[0112]

[0113] Among them, Q is the objective function value in the process of scene reduction by K-medoids, M is the initial real-time scene set, N is the optimal subset of M, and u i are any two scenarios in the wind power scenario set, u j For any two scenes in the photovoltaic scene set, d(u i ,u j ) are two scenes u i 、u j The distance between them, T is the number of scheduling period periods, and pi is the probability of scene i appearing.

[0114] The basic steps of K-medoids scene reduction are:

[0115] a) Randomly select r scenes from the initial real-time scene set as the initial clustering centers. express.

[0116] b) Based on the principle of closest distance to the center, Distribute the remaining objects into various categories.

[0117] c) According to the objective function Based on the minimum principle, new cluster centers are found to replace the original cluster centers.

[0118] d) Determine whether it converges. If not, repeat step b). If it converges, the r cluster centers {R1, R2, …, R r} is the r representative scenes after scene reduction.

[0119] Calculate the probability of occurrence of the scene after reduction, that is, π(r) is the ratio of the number of scenes in the rth category to the total number of scenes.

[0120] The wind power generation scenario S and the photovoltaic power generation scenario Y are reduced to the wind power scenario S and the photovoltaic scenario Y respectively, and the two are combined to form n = s × y scenarios, which is the final real-time scenario set.

[0121] Furthermore, in step S6, the second stage objective function is:

[0122]

[0123] Among them, F2 is the objective function value of the second stage, n is the real-time scene number, π(n) is the probability of real-time scene n, t is the time, The price increase for conventional units in real time is The price reduction for conventional units in real time is The price increase for the photovoltaic power station p in real time, The price reduction for the photovoltaic power station p in real time, is the real-time price increase of wind turbine w, The price reduction of wind turbine w in real time, is the power increase of conventional unit g in real-time stage under real-time scenario n, is the power reduction of conventional unit g in real-time scenario n, is the power increase of PV power station p in real-time scenario n, is the power reduction of PV power station p in real-time scenario n, is the power increase of wind turbine w in real-time scenario n, is the power reduction of wind turbine w in real-time scenario n, c pv is the cost coefficient of abandoned light in the real-time stage, c wind The wind curtailment cost coefficient in real time, c load is the load loss penalty cost coefficient, is the amount of abandoned light from PV power station p in real-time scenario n, is the abandoned wind volume of wind turbine w in real-time scenario n, is the load loss of node j at time t in real-time scenario n.

[0124] The operating constraints of the second phase include:

[0125] The second system power balance constraints, conventional unit re-dispatching reserve capacity constraints, photovoltaic power stations and wind turbine reserve capacity constraints, and electric vehicle charging and discharging power and charge state constraints in real-time scenarios.

[0126] The second system power balance constraint is:

[0127]

[0128] Among them, P g,t,n is the output power of conventional unit g at time t in real-time scenario n, P w,t,n is the output of the wind turbine at time t in real-time scenario n, P p,t,n is the output of photovoltaic power station p at time t in real-time scenario n, is the charging and discharging power of fast charging station cluster k at time t in real-time scenario n, L j,t is the load forecast value of node j in the distribution network at time t, is the load loss of node j at time t in real-time scenario n.

[0129] The reserve capacity constraint for re-dispatching of conventional units is:

[0130]

[0131]

[0132]

[0133]

[0134] in, is the power increase of conventional unit g in real-time stage under real-time scenario n, is the power reduction of conventional unit g in real-time scenario n, is the reserve capacity of conventional unit g at time t, is the reserve capacity of conventional unit g at time t, P g,t is the output power of conventional unit g at time t, is the flag bit of the power increase of conventional unit g at time t in real-time scenario n, is the flag bit of the power reduction of conventional unit g at time t in real-time scenario n, and It is a 0-1 variable, 1 represents regulated power, and 0 represents unregulated power.

[0135] The reserve capacity constraints of photovoltaic power stations and wind turbines are:

[0136]

[0137]

[0138]

[0139]

[0140]

[0141]

[0142]

[0143] in, is the predicted output value of wind turbine w at time t, is the predicted output value of the photovoltaic power station p at time t, is the maximum reserve capacity of the photovoltaic power station p at time t, is the maximum reserve capacity of the photovoltaic power station p at time t, is the maximum reserve capacity of wind turbine w at time t, is the maximum reserve capacity of wind turbine w at time t, is the flag for wind turbine w to increase power at time t in real-time scenario n, is the flag bit of wind turbine w reducing power at time t in real-time scenario n, It is the flag bit of the photovoltaic power station p increasing power at time t in the real-time scenario n. It is the flag bit of the photovoltaic power station p reducing the power at time t in the real-time scenario n. and It is a 0-1 variable, 1 represents regulated power, and 0 represents unregulated power.

[0144] The charging and discharging power and charge state constraints of electric vehicles in real-time scenarios are:

[0145]

[0146]

[0147]

[0148] in, is the power of fast charging station cluster k at time t in real-time scenario n, is the state of charge of fast charging station cluster k at time t in real-time scenario n, is the maximum power of fast charging cluster k at time t, is the minimum power of fast charging cluster k at time t, is the maximum state of charge of fast charging cluster k at time t, is the minimum state of charge of fast charging cluster k at time t, and Δt is the time interval between time t and time t+1.

[0149] Furthermore, in step S7, the objective function of the two-stage stochastic optimization scheduling model is the sum of the objective function of the first stage and the objective function of the second stage:

[0150]

[0151] The constraints in the first-stage day-ahead scheduling and the second-stage real-time scheduling are combined to form the constraints of the two-stage stochastic optimization scheduling model.

[0152] The present invention uses historical data of wind power and photovoltaic power to generate uncertain wind power and photovoltaic power scenarios that conform to the spatiotemporal characteristics based on a generative adversarial network, and represents the uncertainty of electric vehicle charging demand with an electric vehicle fast charging station cluster model. Finally, a two-stage random optimization scheduling model for the power system that considers the uncertainty of wind, solar, and electric vehicles is constructed to determine the unit combination and backup plan, promote the consumption of renewable energy, improve the economic benefits of the power grid, and help achieve the goals of "carbon peak" and "carbon neutrality". The present invention takes into account the uncertainty of the output of new energy such as wind and solar power and the charging demand of electric vehicles, and realizes the coordinated regulation of new energy, electric vehicle clusters, and traditional units, which is conducive to the operation efficiency of the power grid, alleviates the imbalance between supply and demand of transportation, and enhances user experience, with good economic and environmental benefits.

[0153] like Figure 3 As shown, the distribution network dispatching system considering the uncertainty of wind, solar and electric vehicles of the present invention includes: a cluster constraint determination unit 1, a day-ahead forecast data acquisition unit 2, a first target determination unit 3, a first constraint determination unit 4, a real-time scenario generation unit 5, a second target and constraint determination unit 6, a model building unit 7 and a solving unit 8.

[0154] The cluster constraint determination unit 1 is used to determine the fast charging station cluster constraint according to the charging and discharging conditions of each electric vehicle in the fast charging station cluster at a historical moment.

[0155] The day-ahead forecast data acquisition unit 2 is used to acquire the day-ahead forecast data of conventional units in the distribution network, the day-ahead forecast data of wind turbines, the day-ahead forecast data of photovoltaic power stations and the load forecast value of each node.

[0156] The first target determination unit 3 is connected to the day-ahead forecast data acquisition unit 2, and the first target determination unit 3 is used to determine the first stage objective function according to the day-ahead forecast data of the conventional unit.

[0157] The first constraint determination unit 4 is connected to the day-ahead forecast data acquisition unit 2, and is used to determine the first-stage operation constraint conditions based on the day-ahead forecast data of conventional units, the day-ahead forecast data of wind turbines, the day-ahead forecast data of photovoltaic power stations, and the load forecast values ​​of each node.

[0158] The real-time scenario generation unit 5 is connected to the day-ahead forecast data acquisition unit 2, and is used to generate a real-time scenario set based on the day-ahead forecast data of the wind turbine and the day-ahead forecast data of the photovoltaic power station by using a generative adversarial network. The real-time scenario set includes multiple real-time scenarios, each of which includes the active output of the wind turbine and the active output of the photovoltaic power station.

[0159] The second target and constraint determination unit 6 is connected to the real-time scenario generation unit 5, and is used to determine the second-stage target function and the second-stage operation constraint conditions according to the real-time scenario set.

[0160] The model building unit 7 is connected to the cluster constraint determination unit 1, the first target determination unit 3, the first constraint determination unit 4 and the second target and constraint determination unit 6. The model building unit 7 is used to establish a two-stage stochastic optimization scheduling model according to the first-stage objective function, the second-stage objective function, the fast charging station cluster constraints, the first-stage operating constraints and the second-stage operating constraints.

[0161] The solving unit 8 is connected to the model building unit 7, and the solving unit 8 is used to solve the two-stage stochastic optimization scheduling model to obtain the scheduling result of the distribution network.

[0162] Compared with the prior art, the distribution network dispatching system of the present invention that takes into account the uncertainties of wind, solar and electric vehicles has the same beneficial effects as the above-mentioned distribution network dispatching method that takes into account the uncertainties of wind, solar and electric vehicles, which will not be repeated here.

[0163] In this specification, each embodiment is described in a progressive manner, and each embodiment focuses on the differences from other embodiments. The same or similar parts between the embodiments can be referred to each other. For the system disclosed in the embodiment, since it corresponds to the method disclosed in the embodiment, the description is relatively simple, and the relevant parts can be referred to the method part.

[0164] The principles and implementation methods of the present invention are described in this article using specific examples. The description of the above embodiments is only used to help understand the method and core idea of ​​the present invention. At the same time, for those skilled in the art, according to the idea of ​​the present invention, there will be changes in the specific implementation methods and application scope. In summary, the content of this specification should not be understood as limiting the present invention.

Claims

1. A distribution network dispatching method considering the uncertainty of wind, solar and electric vehicles, characterized in that: The distribution network dispatching method considering the uncertainty of wind, solar and electric vehicles includes: According to the charging and discharging conditions of each electric vehicle in the fast charging station cluster at a historical moment, the fast charging station cluster constraints are determined; Obtain the day-ahead forecast data of conventional units in the distribution network, the day-ahead forecast data of wind turbines, the day-ahead forecast data of photovoltaic power stations, and the load forecast value of each node; According to the day-ahead forecast data of conventional units, the first-stage objective function is determined; the first-stage objective function is to minimize the startup cost, shutdown cost, power generation cost, increase reserve cost, decrease reserve cost and power generation cost; Determine the operating constraints of the first phase based on the day-ahead forecast data of conventional units, the day-ahead forecast data of wind turbines, the day-ahead forecast data of photovoltaic power stations, and the load forecast values ​​of each node; Based on the day-ahead forecast data of the wind turbine and the day-ahead forecast data of the photovoltaic power station, a generative adversarial network is used to generate a real-time scenario set; the real-time scenario set includes multiple real-time scenarios, each of which includes the active output of the wind turbine and the active output of the photovoltaic power station; According to the real-time scenario set, the second-stage objective function and the second-stage operation constraints are determined; the second-stage objective function is: Among them, F2 is the objective function value of the second stage, n is the real-time scene number, π(n) is the probability of real-time scene n, t is the time, The price increase for conventional units in real time is The price reduction for conventional units in real time is The price increase for the photovoltaic power station p in real time, The price reduction for the photovoltaic power station p in real time, is the real-time price increase of wind turbine w, The price reduction of wind turbine w in real time, is the power increase of conventional unit g in real-time stage under real-time scenario n, is the power reduction of conventional unit g in real-time scenario n, is the power increase of PV power station p in real-time scenario n, is the power reduction of PV power station p in real-time scenario n, is the power increase of wind turbine w in real-time scenario n, is the power reduction of wind turbine w in real-time scenario n, c pv is the cost coefficient of abandoned light in the real-time stage, c wind The wind curtailment cost coefficient in real time, c load is the load loss penalty cost coefficient, is the amount of abandoned light of photovoltaic power station p at time t in real-time scenario n, is the wind abandonment volume of wind turbine w at time t in real-time scenario n, is the load loss of node j at time t in real-time scenario n; Establishing a two-stage stochastic optimization scheduling model according to the first-stage objective function, the second-stage objective function, the fast charging station cluster constraint, the first-stage operation constraint condition and the second-stage operation constraint condition; The two-stage random optimization scheduling model is solved to obtain the scheduling results of the distribution network; the scheduling results include the combination of conventional units, the reserve capacity plan, the output of conventional units, the power adjustment of wind turbines and the power adjustment of photovoltaic power stations.

2. The distribution network dispatching method considering the uncertainty of wind, solar and electric vehicles according to claim 1 is characterized in that: The fast charging station cluster constraints are determined according to the charging and discharging conditions of each electric vehicle in the fast charging station cluster at a historical moment, specifically including: For any electric vehicle in the fast charging station cluster, determine the maximum charge state of the electric vehicle according to the initial power of the electric vehicle when it is connected to the fast charging station cluster, the maximum charge and discharge power of the electric vehicle, and the maximum power that the electric vehicle can reach; Determine the minimum state of charge of the electric vehicle according to the initial power of the electric vehicle when it is connected to the fast charging station cluster, the minimum charge and discharge power of the electric vehicle, the time when the electric vehicle leaves the fast charging station cluster, the expected state of charge of the electric vehicle when it leaves the fast charging station cluster, the maximum charge and discharge power of the electric vehicle and the minimum power that the electric vehicle can reach; Determining the maximum state of charge of the fast charging station cluster according to the maximum state of charge of each electric vehicle in the fast charging station cluster; Determining the minimum state of charge of the fast charging station cluster according to the minimum state of charge of each electric vehicle in the fast charging station cluster; Determining the maximum power of the fast charging station cluster according to the power upper bound of each electric vehicle in the fast charging station cluster; Determining the minimum power of the fast charging station cluster according to the lower bound of the power of each electric vehicle in the fast charging station cluster; A fast charging station cluster constraint is determined based on a maximum state of charge of the fast charging station cluster, a minimum state of charge of the fast charging station cluster, a maximum power of the fast charging station cluster, and a minimum power of the fast charging station cluster.

3. The distribution network dispatching method considering the uncertainty of wind, solar and electric vehicles according to claim 1 is characterized in that: The day-ahead forecast data of the conventional units include: startup cost, shutdown cost, power generation cost, reserve price increase, reserve price decrease, reserve quantity increase, reserve quantity decrease; The following formula is used to determine the first-stage objective function: f g (P g,t )=a g (P g,t ) 2 +b g P g,t +c g ; Among them, F1 is the objective function value of the first stage, t is the time, g represents the conventional unit, is the startup cost of conventional unit g, is the downtime cost of conventional unit g, u g,t is a binary variable, u g,t The value of 1 indicates that the conventional unit g is in the startup process at time t, u g,t A value of 0 indicates that the conventional unit g is not in the startup process at time t. g,t is a binary variable, v g,t The value of 1 indicates that the conventional unit g is in the shutdown process at time t, v g,t A value of 0 indicates that the conventional unit g is not in the shutdown process at time t. g (P g,t ) is the power generation cost of conventional unit g, P g,t is the output power of conventional unit g at time t, a g 、b g 、c g is the power generation cost coefficient of conventional unit g, The reserve price of conventional unit g is increased. The reserve price of conventional unit g is lowered. is the reserve capacity of conventional unit g at time t, It is the reserve capacity of conventional unit g at time t.

4. The distribution network dispatching method considering the uncertainty of wind, solar and electric vehicles according to claim 1 is characterized in that: The operating constraints of the first stage include the first system power balance constraints, conventional unit output and spare capacity constraints, conventional unit start and stop constraints, conventional unit start and stop time constraints, conventional unit ramp constraints and network security constraints.

5. The distribution network dispatching method considering the uncertainty of wind, solar and electric vehicles according to claim 4 is characterized in that: The day-ahead prediction data of the conventional unit includes the output power of the conventional unit; the day-ahead prediction data of the wind turbine unit includes the output prediction value of the wind turbine unit; the day-ahead prediction data of the photovoltaic power station includes the output prediction value of the photovoltaic power station; The first system power balance constraint is: Among them, g represents conventional units, w represents wind turbines, p represents photovoltaic power stations, j represents nodes in the distribution network, k represents charging station clusters, and P g,t is the output power of conventional unit g at time t, is the predicted output value of wind turbine w at time t, is the output prediction value of the photovoltaic power station p at time t, L j,t is the load forecast value of node j in the distribution network at time t, is the overall power of charging station cluster k at time t.

6. The distribution network dispatching method considering the uncertainty of wind, solar and electric vehicles according to claim 4 is characterized in that: The day-ahead forecast data of the conventional unit includes the operating status, output upper limit, output lower limit, output power, upward reserve capacity, downward reserve capacity, maximum downward reserve capacity and maximum upward reserve capacity of the conventional unit; The conventional unit output and reserve capacity constraints are: Among them, i g,t is the operating status of the conventional unit g at time t, is the output upper limit of conventional unit g, is the lower limit of the output of the conventional unit g, P g,t is the output power of conventional unit g at time t, is the reserve capacity of conventional unit g at time t, is the reserve capacity of conventional unit g at time t, is the maximum down-regulated reserve capacity of conventional unit g at time t, It is the maximum upward reserve capacity of conventional unit g at time t.

7. The distribution network dispatching method considering the uncertainty of wind, solar and electric vehicles according to claim 1 is characterized in that: The real-time scenario set is generated by using a generative adversarial network based on the day-ahead forecast data of the wind turbine and the day-ahead forecast data of the photovoltaic power station, specifically including: Based on the day-ahead forecast data of wind turbines and photovoltaic power plants, a generative adversarial network is used to generate a preliminary real-time scenario set. The K-medoids method is used to reduce the preliminary real-time scene set to obtain the final real-time scene set.

8. The distribution network dispatching method considering the uncertainty of wind, solar and electric vehicles according to claim 1 is characterized in that: The second stage operation constraints include: The second system power balance constraints, conventional unit re-dispatching reserve capacity constraints, photovoltaic power stations and wind turbine reserve capacity constraints, and electric vehicle charging and discharging power and charge state constraints in real-time scenarios.

9. A distribution network dispatching system considering the uncertainty of wind, solar and electric vehicles, characterized in that: The distribution network dispatching system considering the uncertainty of wind, solar and electric vehicles includes: A cluster constraint determination unit, used to determine the fast charging station cluster constraint according to the charging and discharging conditions of each electric vehicle in the fast charging station cluster at a historical moment; A day-ahead forecast data acquisition unit is used to acquire the day-ahead forecast data of conventional units in the distribution network, the day-ahead forecast data of wind turbines, the day-ahead forecast data of photovoltaic power stations, and the load forecast value of each node; A first target determination unit is connected to the day-ahead forecast data acquisition unit, and is used to determine a first-stage target function according to the day-ahead forecast data of the conventional unit; the first-stage target function is to minimize the startup cost, the shutdown cost, the power generation cost, the increase in the reserve cost, the decrease in the reserve cost, and the power generation cost; a first constraint determination unit connected to the day-ahead forecast data acquisition unit, and configured to determine the first-stage operation constraint conditions according to the day-ahead forecast data of the conventional unit, the day-ahead forecast data of the wind turbine unit, the day-ahead forecast data of the photovoltaic power station, and the load forecast value of each node; A real-time scenario generation unit is connected to the day-ahead prediction data acquisition unit, and is used to generate a real-time scenario set based on the day-ahead prediction data of the wind turbine generator set and the day-ahead prediction data of the photovoltaic power station by using a generative adversarial network; the real-time scenario set includes a plurality of real-time scenarios, each of which includes the active output of the wind turbine generator set and the active output of the photovoltaic power station; The second objective and constraint determination unit is connected to the real-time scenario generation unit and is used to determine the second-stage objective function and the second-stage operation constraint conditions according to the real-time scenario set; the second-stage objective function is: Among them, F2 is the objective function value of the second stage, n is the real-time scene number, π(n) is the probability of real-time scene n, t is the time, The price increase for conventional units in real time is The price reduction for conventional units in real time is The price increase for the photovoltaic power station p in real time, The price reduction for the photovoltaic power station p in real time, is the real-time price increase of wind turbine w, The price reduction of wind turbine w in real time, is the power increase of conventional unit g in real-time stage under real-time scenario n, is the power reduction of conventional unit g in real-time scenario n, is the power increase of PV power station p in real-time scenario n, is the power reduction of PV power station p in real-time scenario n, is the power increase of wind turbine w in real-time scenario n, is the power reduction of wind turbine w in real-time scenario n, c pv is the cost coefficient of abandoned light in the real-time stage, c wind The wind curtailment cost coefficient in real time, c load is the load loss penalty cost coefficient, is the amount of abandoned light of photovoltaic power station p at time t in real-time scenario n, is the wind abandonment volume of wind turbine w at time t in real-time scenario n, is the load loss of node j at time t in real-time scenario n; a model building unit connected to the cluster constraint determination unit, the first target determination unit, the first constraint determination unit, and the second target and constraint determination unit, and configured to establish a two-stage stochastic optimization scheduling model according to the first-stage objective function, the second-stage objective function, the fast charging station cluster constraint, the first-stage operation constraint condition, and the second-stage operation constraint condition; A solving unit is connected to the model to solve the two-stage random optimization scheduling model to obtain a scheduling result of the distribution network.

Images