Optimal scheduling method for distribution network-cloud energy storage system considering wind and light uncertainty

Abstract

The present application relates to the technical field of multi-energy interconnection system scheduling, and discloses a power distribution network-cloud energy storage system optimal scheduling method considering wind and light uncertainty. By co-modeling the power distribution network, the cloud energy storage system and the cloud energy storage user, a deterministic model and an uncertainty model of the power distribution network-cloud energy storage system double-layer optimal scheduling considering wind and light uncertainty are established, the Monte Carlo simulation method and the synchronous back substitution method are used to generate wind and light generator set output uncertainty output scene, and the Gurobi solver is used for solving. In the power distribution network optimal scheduling, the cloud energy storage operation mode is considered, and the user side dispersed energy storage resources can be fully utilized. When the power distribution network formulates the day-ahead scheduling plan, only information interaction with the cloud energy storage operator is needed, and the management difficulty can be greatly reduced.

Description

Technical Field

[0001] This invention relates to the field of multi-energy interconnected system scheduling technology, and in particular to an optimized scheduling method for a distribution network-cloud energy storage system that takes into account the uncertainties of wind and solar power. Background Technology

[0002] With the increasing proportion of new energy sources connected to the distribution network and the advancement of energy storage technology, cloud energy storage may become a new approach to user-side energy storage management in the future. The distribution network user side has a large amount of decentralized energy storage resources with high idle rates and significant management challenges. Rational utilization of these resources can not only promote the absorption of new energy sources and reduce the operating costs of the distribution network, but also improve the utilization rate of idle decentralized energy storage resources. Existing research typically treats a certain scale of energy storage as a unified whole through clustering effects for unified management. Besides the aforementioned centralized energy storage, the potential for utilization of user-side decentralized energy storage resources such as electric vehicles, user-built energy storage, and demand response loads is also very large. Rational utilization of these resources can not only promote the absorption of new energy sources and reduce the operating costs of the distribution network, but also improve the utilization rate of idle decentralized energy storage resources. Therefore, applying cloud energy storage to the optimal scheduling of the distribution network and researching a two-layer optimal scheduling method for the distribution network-cloud energy storage system that considers the uncertainties of wind and solar power is of great significance. Summary of the Invention

[0003] To address the aforementioned problems, the present invention aims to provide an optimized scheduling method for a distribution network-cloud energy storage system that considers the uncertainties of wind and solar power. This method involves collaboratively modeling gas turbines, wind and solar power generators, substations, cloud energy storage systems, and participating electric vehicle users, uninterruptible power supply users, and demand response load users within the distribution network. During the optimized scheduling process, the distribution network only needs to interact with the cloud energy storage system. The distribution network manages the charging and discharging of the cloud energy storage system, while the cloud energy storage operator manages the charging and discharging of cloud energy storage users. The solution is obtained using the Gurobi solver. The technical solution is as follows:

[0004] An optimized scheduling method for a distribution network-cloud energy storage system that takes into account the uncertainties of wind and solar power includes the following steps:

[0005] Step 1: Identify the various components of the distribution network-cloud energy storage system, including the distribution network, the cloud energy storage system, and cloud energy storage users. Cloud energy storage users include electric vehicle users, uninterruptible power supply users, and demand response load users.

[0006] Step 2: Establish a deterministic model for the two-layer optimal scheduling of the distribution network-cloud energy storage system, taking into account the uncertainties of wind and solar power.

[0007] The deterministic model includes an upper-level model and a lower-level model. The constraints of the upper-level model include constraints on cloud energy storage systems, constraints on distribution network units, and constraints on power balance. The constraints of the lower-level model include constraints on cloud energy storage users and constraints on distribution network power flow. The constraints on cloud energy storage users include constraints on electric vehicles, constraints on uninterruptible power supplies, and constraints on demand response loads.

[0008] Step 3: Establish an uncertainty model for the two-layer optimal scheduling of the distribution network-cloud energy storage system, taking into account the uncertainties of wind and solar power.

[0009] When considering the uncertainty of wind and solar power output, typical wind and solar power output scenarios are incorporated into the upper-level model. Combined with the uncertainty model, a scheduling strategy that considers the uncertainty of wind and solar power output is obtained, and the scheduling strategy ensures that the distribution network can still operate stably when wind and solar power output fluctuates.

[0010] Step 4: Under the premise of satisfying the safety constraints of each component of the system, establish a two-layer optimization scheduling model of the distribution network-cloud energy storage system that takes into account the uncertainties of wind and solar power. The objective function of the upper-layer model is to minimize the operating cost of the distribution network, and the optimal output plan of each power source in the distribution network and the charging and discharging plan of the cloud energy storage system are obtained by solving the model. The objective function of the lower-layer model is to maximize the profit of the cloud energy storage operator. Based on the charging and discharging plan of the cloud energy storage system obtained by the upper-layer model, and combined with the operating constraints of the distribution network, the charging and discharging plans of each electric vehicle user, uninterruptible power supply user, and demand response load user are formulated.

[0011] Step 5: Use Monte Carlo simulation to generate uncertain power output scenarios for wind and solar generators, then use synchronous back-substitution method to reduce the scenarios, and finally generate typical wind and solar power output scenarios.

[0012] Step 6: Input the operating parameters of the distribution network-cloud energy storage system, and use a commercial solver to solve the two-layer optimal scheduling model of the distribution network-cloud energy storage system that takes into account the uncertainties of wind and solar power, so as to obtain its optimal scheduling strategy.

[0013] Furthermore, in step 1, the power distribution network includes substation units, gas turbine units, wind and solar generator units, cloud energy storage systems, cloud energy storage users, electrical loads, bus nodes, and transmission lines.

[0014] Furthermore, the constraints of the cloud energy storage system, distribution network unit, and power balance in the upper-level model in step 2 are specifically as follows:

[0015] 1) Constraints of cloud energy storage systems

[0016] The constraints of the cloud energy storage system include the charging and discharging power and capacity constraints of the cloud energy storage system. The cloud energy storage operator determines the upper and lower limits of the charging and discharging power of the cloud energy storage system in each time period and the upper and lower limits of the net charging power on the day by estimating the capacity limit of the cloud energy storage users under its management that can be dispatched in each time period.

[0017]

[0018]

[0019]

[0020]

[0021]

[0022]

[0023] In the formula: t is the time index, T is the total scheduling period; e is the index of the cloud energy storage system, and E is the set of cloud energy storage systems in the distribution network; and Let t represent the discharge power and charging power of the cloud energy storage system e during time period t. and These represent the upper limits of discharge power and charging power of cloud energy storage system e during time period t, respectively. and These represent the lower limits of discharge power and charging power of cloud energy storage system e during time period t, respectively. The net charge of cloud energy storage system e. and These represent the upper and lower limits of the net charging capacity of the cloud energy storage system; O e,t Let t be the capacity of cloud energy storage system e during time period t. and These are the upper and lower limits of the capacity of the cloud energy storage system, respectively.

[0024] 2) Constraints of distribution network units

[0025] The constraints on the power distribution network units include the operating constraints of gas turbine units, the output constraints of wind and solar turbine units, and the transmission power constraints of substations.

[0026] a) Operating constraints of gas turbine units

[0027] The operational constraints of gas turbine units specifically include upper and lower limits for output, minimum start-up time constraints, minimum shutdown time constraints, start-up and shutdown fuel consumption constraints, and ramp-up constraints; as shown in the following formulas:

[0028]

[0029]

[0030]

[0031]

[0032] K g,t ≥k g ·(x g,t -x g,t-1 ),K g,t ≥0, g∈G, t∈T

[0033] D g,t ≥d g ·(x g,t-1 -x g,t ),D g,t ≥0, g∈G, t∈T

[0034]

[0035]

[0036] In the formula: g is the index of the gas turbine unit, G is the set of gas turbine units in the power distribution network; x g,t P represents the operating state of the gas turbine unit g during time period t, where 1 indicates it is in the start-up state and 0 indicates it is in the stop-down state; g,t and Q g,t These represent the active power and reactive power output by the gas turbine unit g during time period t, respectively. and These are the maximum and minimum limits for the active power output of the gas turbine unit, respectively. and These are the maximum and minimum limits for the reactive power output of the gas turbine unit, respectively. and These are the start-up and shutdown time counters for gas turbine unit g during time period t; and These represent the minimum start-up time and minimum shutdown time for the gas turbine unit, respectively; K g,t and D g,t The fuel consumption of the gas turbine unit during time period t for starting and stopping (g and k) are respectively. g and d g These represent the fuel consumption for the gas turbine unit g during startup and shutdown, respectively. and These are the uphill and downhill ramp rates for gas turbine unit g, respectively.

[0037] b) Output constraints of wind and solar turbine units

[0038] The output constraints of wind power and photovoltaic units include active power output constraints and reactive power output constraints;

[0039]

[0040]

[0041]

[0042]

[0043] In the formula: S and W are the sets of photovoltaic generators and wind turbine generators in the distribution network, respectively. and P represents the predicted power output of photovoltaic generator s and wind turbine w during time period t. s,t and P w,t Q represents the actual output values ​​of photovoltaic generator s and wind turbine w during time period t. s,t and Q w,t Let t represent the reactive power output of photovoltaic generator s and wind turbine generator w during time period t. and The power factor angles of photovoltaic generator s and wind turbine generator w are respectively.

[0044] c) Substation transmission power constraints

[0045] Substation transmission power constraints include active power limit constraints and reactive power limit constraints;

[0046]

[0047]

[0048] In the formula: Z represents the set of substations in the distribution network. The power purchased by the distribution network from the upper-level power grid through substation z during time period t. The reactive power supplied by the upstream power grid to the distribution network through substation z during time period t. and These represent the upper and lower limits of active power transmitted from the upstream power grid to the distribution network through substation z, respectively. and These are the upper and lower limits of reactive power transmitted from the upper-level power grid to the distribution network through substation z, respectively.

[0049] 3) Power balance constraints

[0050] Power balance constraints include active power balance constraints and reactive power balance constraints:

[0051]

[0052]

[0053] In the formula, D represents the set of nodes where the load is located; P t loss Let C be the estimated network loss power of the distribution network in time period t in the upper-level model; let C be the set of reactive power compensation devices, and Q be the estimated network loss power of the distribution network in time period t. c,t P represents the reactive power generated by reactive power compensation device c during time period t; d,t and Q d,t Let d be the active power demand and reactive power demand of load d during time period t.

[0054] Furthermore, the cloud energy storage user constraints and distribution network power flow constraints in the lower-level model in step 2 are specifically as follows:

[0055] 1) Cloud energy storage user constraints

[0056] Cloud energy storage user constraints include electric vehicle constraints, demand response load constraints, and uninterruptible power supply constraints.

[0057] a) Electric vehicle constraints

[0058] Electric vehicle constraints include electric vehicle charging and discharging capacity constraints, electric vehicle charging and discharging power constraints, electric vehicle minimum charging and discharging time constraints, and electric vehicle charging station total charging and discharging power limit constraints.

[0059]

[0060]

[0061]

[0062]

[0063]

[0064]

[0065]

[0066]

[0067]

[0068]

[0069]

[0070]

[0071] In the formula: V represents the set of electric vehicles; and These represent the charging power and discharging power of electric vehicle v during time period t; O v,t Let η be the capacity of electric vehicle v during time period t. v Improve the charging and discharging efficiency of electric vehicles; and These represent the upper and lower limits of the capacity of electric vehicle v, respectively. This refers to the initial battery level of electric vehicle V when it is connected to the grid. The expected charge level of electric vehicle v when it leaves the charging station; The net charge amount for electric vehicle v; and These represent the upper and lower limits of the discharge power (v) of the electric vehicle, respectively. Let v be the discharge state of electric vehicle v during time period t, where 1 indicates discharge and 0 indicates no discharge; and These represent the upper and lower limits of the charging power of electric vehicle V, respectively. The charging status of electric vehicle v during time period t is represented by 1, indicating charging and 0, indicating no charging. A counter for the offline time of an electric vehicle after it has been fully charged during time period t. Minimum off-grid time after electric vehicle is fully charged; This is a counter for the duration of time an electric vehicle remains off-grid after it has fully discharged during time period t. The minimum off-grid time for an electric vehicle after it has completely discharged its power; This is a counter for the on-grid charging time of electric vehicles during period t. Minimum charging time for electric vehicles on the grid; This is a counter for the on-grid discharge duration of electric vehicles during time period t. Let A(y) be the minimum discharge time of an electric vehicle on the grid; let A(y) be the set of equipment within electric vehicle charging station y, and let Y be the set of electric vehicle charging stations. and These represent the maximum charging power and discharging power of electric vehicle charging station y, respectively.

[0072] b) Demand response load constraints

[0073] Demand response loads include interruptible loads, and the constraint is the interruptible load limit constraint.

[0074]

[0075] In the formula: Let d be the interruptible load amount during time period t. α represents the participation status of interruptible load d during time period t; a value of 1 indicates it is in an interrupted state, and a value of 0 indicates it is not interrupted. d This refers to the proportion of interruptible load capacity to total load.

[0076] c) Uninterruptible power supply constraints

[0077] Uninterruptible power supply (UPS) constraints include UPS capacity constraints and UPS charging and discharging power constraints.

[0078]

[0079]

[0080] O u,T =O u,0

[0081]

[0082]

[0083]

[0084] In the formula: U is the set of uninterruptible power supplies; and These represent the charging power and subsequent discharging power of the uninterruptible power supply u during time period t; O u,t Let η be the capacity of the uninterruptible power supply u during time period t. u For uninterruptible power supply charging and discharging efficiency; and The upper and lower limits of the uninterruptible power supply capacity (U); O u,T and O u,0 These represent the capacities of the uninterruptible power supply u at the end and beginning of the scheduling cycle, respectively. and These represent the upper and lower limits of the discharge power of the uninterruptible power supply (UPS), respectively. The value represents the discharge state of the uninterruptible power supply u during time period t. A value of 1 indicates discharge, and a value of 0 indicates no discharge. and These represent the upper and lower limits of the charging power of the uninterruptible power supply (UPS), respectively. The charging state of the uninterruptible power supply u during time period t is represented by 1, which indicates charging and 0 indicates no charging.

[0085] The sum of the charging and discharging power of each cloud energy storage user in the lower-level model at each time period must be equal to the charging and discharging power of the cloud energy storage system at each time period obtained from the upper-level model.

[0086]

[0087]

[0088]

[0089] In the formula: A(e) is the set of devices in cloud energy storage system e;

[0090] 2) Distribution network power flow constraints

[0091] A second-order cone relaxation method is used to linearize the power flow constraints of the distribution network. The operational constraints of the distribution network include power balance constraints at distribution network nodes, voltage drop limits on distribution lines, second-order cone constraints, and line operation constraints. Among these, the difference in network losses between the upper and lower level models is considered. Equal to the network loss value P predicted by the upper-level model t loss Compared with the actual network loss value of the lower-level model The difference;

[0092]

[0093]

[0094]

[0095]

[0096]

[0097]

[0098]

[0099]

[0100]

[0101] In the formula: A(i) is the set of distribution network equipment directly connected to node i; L and N bus Let be the set of transmission lines and nodes in the distribution network, respectively; s(l) and r(l) are the sending-end node and receiving-end node of transmission line l; P l,t and Q l,t These represent the active power flow and reactive power flow of transmission line l during time period t, respectively. and The active and reactive power losses of transmission line l during time period t; ω represents the difference in network loss between the upper and lower layer models. l,t Let ψ be the square of the current in transmission line l during time period t. j,t and ψ i,t Let r be the squared voltage values ​​of node j and node i respectively during time period t; l and x l These are the resistance and reactance values ​​of transmission line l, respectively. and These represent the squares of the upper and lower limits of the allowable current to flow through transmission line l, respectively. and These represent the squares of the upper and lower voltage limits that node i can withstand, respectively.

[0102] Furthermore, in step 3, the uncertainty model for wind and solar power output is considered as follows:

[0103]

[0104]

[0105]

[0106]

[0107]

[0108]

[0109]

[0110]

[0111]

[0112]

[0113]

[0114]

[0115]

[0116]

[0117]

[0118]

[0119]

[0120]

[0121] In the formula: the superscript sce indicates the value of the variable under scenario sce, and γ represents the set of typical wind and solar power output scenarios; For the emergency ramp rate of the gas turbine unit; This refers to the adjustable range of substation transmission power under uncertain scenarios.

[0122] Furthermore, in step 4, the two-layer optimization scheduling model of the distribution network-cloud energy storage system...

[0123] (1) In the upper-level model, the distribution network operation cost in the objective function includes the cost of using the cloud energy storage system, the cost of purchasing electricity from the upper-level grid, the cost of generating electricity from gas turbine units, and the penalty for renewable energy curtailment. The objective function expression is as follows:

[0124] minF1=F M +F E +F G +F Voll

[0125] In the formula: F1 is the operating cost of the distribution network, F M F represents the cost of electricity purchased by the distribution network from the upstream power grid. E For the cost of using cloud energy storage systems, F G For the cost of generating electricity from gas turbine units, F Voll The cost of curtailing renewable energy;

[0126] in:

[0127]

[0128] In the formula: The electricity purchase price corresponding to time period t;

[0129] When the distribution network needs the cloud energy storage system to charge / discharge, it pays the corresponding service fee to the cloud energy storage operator, while the cloud energy storage operator only needs to pay the net charging fee to the distribution network for the net charging amount of the cloud energy storage system.

[0130]

[0131] In the formula: The unit price for charging / discharging services of cloud energy storage system e. The net charging unit price for cloud energy storage system e;

[0132]

[0133] Where: M g (·) represents the heat rate curve. Let be the fuel unit price during time period t;

[0134]

[0135] In the formula: and These are the unit prices for the curtailment penalties for photovoltaic power generation units and wind power generation units, respectively.

[0136] (2) In the lower-level model, the main body is the cloud energy storage system, and the profit of the cloud energy storage operator in the objective function includes the following parts:

[0137] max(F2-F P ) = FE -(F U +F V +F I )-F P

[0138] In the formula: F2 represents the total profit obtained by the cloud energy storage operator, which is equal to the cloud energy storage system usage cost paid by the distribution network to the cloud energy storage operator minus the cloud energy storage user dispatch cost paid by the cloud energy storage operator to the cloud energy storage user. U For uninterruptible power supply (UPS) user dispatch costs, F V For electric vehicle user dispatch costs, F I For interruptible load user scheduling costs, F P The penalty cost is the difference in network loss between upper and lower layer models.

[0139] Cloud energy storage operators pay electric vehicle users for discharge / charge services, and electric vehicle users only need to pay for the net charging portion of the electric vehicle's net charging capacity.

[0140]

[0141] In the formula: The unit price for charging / discharging services for electric vehicles; Net charging cost per unit for electric vehicles;

[0142] Demand response load only considers the interruptible load of users. The cloud energy storage system uses a compensation method to induce users to reduce electricity consumption as part of the discharge power of the cloud energy storage system.

[0143]

[0144] In the formula: Provide cloud energy storage system operators with a compensation price for interruptible loads;

[0145] Cloud energy storage operators pay uninterruptible power supply (UPS) users for discharge / charging services;

[0146]

[0147] In the formula: Unit price for charging / discharging services for uninterruptible power supplies;

[0148] When the upper-level model solves the power output plan of each unit in the distribution network and the charging and discharging plan of the cloud energy storage system, it does not calculate the power flow of the distribution network. Therefore, the network loss in the active power balance constraint is an estimated value. When the lower-level model formulates the charging and discharging plan of each cloud energy storage user, it performs power flow calculation of the distribution network to obtain the actual network loss value. The network loss difference is the difference between the actual network loss and the estimated network loss.

[0149]

[0150] In the formula: c p The penalty unit price is the difference in network loss.

[0151] Furthermore, the process of step 5 is as follows:

[0152] Based on the basic prediction scenarios of wind and solar power generation, and assuming the wind power output prediction error... And photovoltaic power output prediction error It conforms to the standard normal distribution, that is Uncertainty-based power output scenarios were generated using the Monte Carlo simulation method. In scenario h, the power outputs of wind power and solar power are as follows:

[0153]

[0154]

[0155] In the formula: and These represent the predicted power output values ​​of wind turbine w and photovoltaic generator s during time period t, respectively, under scenario sce. and These are the predicted output values ​​of wind turbine w and photovoltaic generator s in time period t, respectively, under the basic prediction scenario; and These represent the output prediction errors of wind turbine w and photovoltaic generator s during time period t, respectively, under scenario sce.

[0156] The scene reduction method is used to select several typical scenes for model solving in order to improve the solution efficiency. The scene reduction steps are as follows:

[0157] Step a: Set SS as the initial scene set; DS is the scene set to be deleted, initially empty; calculate the Euclidean distance between each scene in the initial scene set SS:

[0158]

[0159] In the formula: DT r,k For the r-th scene Ω r and the kth scene Ω k The Euclidean distance between them, P t r and P t k The predicted power output of wind and solar turbines during time period t under scenarios r and k, respectively; Ω u A collection of all scenarios;

[0160] Step b: Calculate the scenario Ω rProbability distance (PD) from all other scenarios k (r), the smaller the value, the higher the similarity between the scene and other scenes;

[0161]

[0162] In the formula: ρ r For scene Ω r The probability of occurrence; r = 1, 2, ..., N, where N is the total number of scenes;

[0163] Step c: Calculate the probability distance between each scene. If there exists a scene Ω... b Probability distance (PD) from all other scenarios k (b) = minPD k (r); then select scene Ω b , as a scenario that will be deleted;

[0164] Step d: Calculate the relationship between each scene and scene Ω in the initial scene set SS. b The Euclidean distance, if DT b,d =minDT b,k If k ≠ b, and b = 1, 2, ..., N, then delete scene Ω. b , will scene Ω b The probability of occurrence is added to scenario Ω d Up; Update the scene set and the probability of each scene occurring:

[0165] SS=SS-{Ω b};DS=DS+{Ω b}

[0166] ρ d =ρ d +ρ b ;ρ b =0

[0167] Step e: Repeat steps bd until the number of scenes meets the target set value.

[0168] Furthermore, the operating parameters of the distribution network-cloud energy storage system include the cost of output of each unit in the distribution network, ramp-up capability, minimum start-up and shutdown time, electrical load size, power transmission limit of transmission lines, current limit of transmission lines, voltage limit of nodes, charging and discharging service fee of cloud energy storage system, capacity limit of cloud energy storage system, charging and discharging power limit of cloud energy storage system, charging and discharging service fee of electric vehicle users, capacity limit of electric vehicle users, charging and discharging power limit of electric vehicle users, charging and discharging service fee of uninterruptible power supply (UPS) users, capacity limit of UPS users, charging and discharging power limit of UPS users, demand response load discharge compensation fee, and demand response load capacity limit.

[0169] The beneficial effects of this invention are as follows:

[0170] To address the issues of high idle rates and management difficulties of distributed energy storage resources on the user side of distribution networks, this invention proposes an optimized scheduling method for distribution network-cloud energy storage systems that considers the uncertainties of wind and solar power generation. By incorporating cloud energy storage operation modes into the optimized scheduling of the distribution network, it can fully utilize distributed energy storage resources on the user side. When formulating day-ahead scheduling plans, the distribution network only needs to interact with the cloud energy storage operator, significantly reducing its management complexity. The optimized scheduling method for distribution network-cloud energy storage systems proposed in this invention incorporates uncertainties in wind and solar power output to address these uncertainties, ensuring stable system operation by adjusting the charging and discharging of the cloud energy storage system during fluctuations in wind and solar power output. Attached Figure Description

[0171] Figure 1 This is a topology diagram of a 33-node distribution network system; GT – Gas turbine; C – Reactive power compensation device; WT – Wind turbine; PV – Photovoltaic turbine; U – Uninterruptible power supply user; EV – Electric vehicle user; D – Interruptible load user; TS – Substation.

[0172] Figure 2(a) shows the power output of photovoltaic units in different scenarios with uncertain wind and solar power output.

[0173] Figure 2(b) shows the wind turbine output under uncertain wind and solar power output scenarios – wind turbine output under different scenarios.

[0174] Figure 3(a) shows the output-optimized scheduling strategy of each unit in Example 1.

[0175] Figure 3(b) shows the output of each unit in Example 1 and the charging and discharging power of the energy storage user.

[0176] Figure 4 This represents the output of each unit in Example 2 – the charging and discharging power of the cloud energy storage system under different scenarios. Detailed Implementation

[0177] To provide a detailed explanation of the technical solutions disclosed in this invention, the invention will be further described below in conjunction with the accompanying drawings and specific embodiments.

[0178] This invention discloses a two-layer optimization scheduling method for a distribution network-cloud energy storage system that takes into account the uncertainties of wind and solar power, comprising the following steps:

[0179] Step 1: Identify the components of the power distribution network-cloud energy storage system.

[0180] Example 1 defines the components of a distribution network-cloud energy storage system: the distribution network, the cloud energy storage system, and cloud energy storage users. The distribution network consists of substation units, gas turbine units, wind and solar generator units, cloud energy storage system, cloud energy storage users, electrical loads, bus nodes, and transmission lines. Cloud energy storage users include electric vehicle users, uninterruptible power supply users, and demand response load users.

[0181] Step 2: Establish a deterministic model for the two-layer optimal scheduling of the power distribution network-cloud energy storage system, taking into account the uncertainties of wind and solar power.

[0182] Example 2 presents a deterministic model for establishing a two-layer optimal scheduling of a distribution network-cloud energy storage system that takes into account the uncertainties of wind and solar power:

[0183] (1) Upper-level model

[0184] The constraints of the upper-level model include two parts: cloud energy storage system constraints and distribution network unit constraints, as well as power balance constraints. The distribution network unit constraints include gas turbine operation constraints, wind and solar turbine output constraints, and substation transmission power constraints.

[0185] The constraints of cloud energy storage systems include charging and discharging power and capacity constraints. Cloud energy storage operators need to estimate the acceptable dispatchable capacity limits of cloud energy storage users under their management in each time period, so as to determine the upper and lower limits of charging and discharging power of the cloud energy storage system in each time period and the upper and lower limits of net charging power on the same day.

[0186]

[0187]

[0188]

[0189]

[0190]

[0191]

[0192] In the formula: t is the time index, T is the total scheduling period, and E is the set of cloud energy storage systems in the distribution network. The discharge and charging power of cloud energy storage system e during time period t. These represent the upper limits of the discharge power and charging power of the cloud energy storage system e during time period t, respectively. These represent the lower limits of the discharge power and charging power of the cloud energy storage system e during time period t, respectively. The net charge of cloud energy storage system e. These represent the upper and lower limits of the net charging capacity of the cloud energy storage system. e,tLet t be the capacity of cloud energy storage system e during time period t. These represent the upper and lower limits of the capacity of the cloud energy storage system.

[0193] The operating constraints of gas turbine units specifically include upper and lower limits for output, minimum start-up time constraints, minimum shutdown time constraints, start-up and shutdown fuel consumption constraints, and ramp-up constraints. As shown in the following formulas:

[0194]

[0195]

[0196]

[0197]

[0198]

[0199]

[0200]

[0201]

[0202] In the formula: G is the set of gas turbine units in the power distribution network, x g,t P represents the operating state of the gas turbine unit g during time period t, where 1 indicates it is in the start-up state and 0 indicates it is in the stop-down state; g,t Q g,t These represent the active and reactive power outputs of the gas turbine unit g during time period t, respectively. These are the maximum and minimum limits of the active power output of the gas turbine unit, respectively. These are the maximum and minimum limits of reactive power output g of the gas turbine unit, respectively. These are the start-up and shutdown time counters for gas turbine unit g, respectively. These represent the minimum start-up time and minimum shutdown time for the gas turbine unit, respectively; K g,t D g,t The fuel consumption of the gas turbine unit during time period t for starting and stopping (g and k) are respectively. g d g These represent the fuel consumption for the gas turbine unit g during startup and shutdown, respectively. These represent the uphill and downhill ramp rates of the gas turbine unit g, respectively.

[0203] The output constraints of wind power and photovoltaic units include active power output constraints and reactive power output constraints.

[0204]

[0205]

[0206]

[0207]

[0208] In the formula: S and W represent the sets of photovoltaic generators and wind turbine generators in the distribution network, respectively. P represents the predicted power output of photovoltaic generator s and wind turbine w during time period t. s,t P w,t Q represents the actual output values ​​of photovoltaic generator s and wind turbine w during time period t. s,t Q w,t Let t represent the reactive power output of photovoltaic generator s and wind turbine generator w during time period t. These are the power factor angles of the photovoltaic generator set s and the wind turbine generator set w, respectively.

[0209] Substation transmission power constraints include active power limit constraints and reactive power limit constraints.

[0210]

[0211]

[0212] In the formula: Z represents the set of substations in the distribution network. The power purchased by the distribution network from the upper-level power grid through substation z during time period t. The reactive power supplied by the upstream power grid to the distribution network through substation z during time period t. These represent the upper and lower limits of active power transmitted from the upstream power grid to the distribution network through substation z, respectively. These represent the upper and lower limits of reactive power transmitted from the upstream power grid to the distribution network through substation z, respectively.

[0213] The power output of each unit in the distribution network, the charging and discharging power of the cloud energy storage system, the load, and the power loss of the network must be balanced during each dispatch period. Power balance constraints include active power balance constraints and reactive power balance constraints.

[0214]

[0215]

[0216] In the formula, D represents the set of nodes where the load is located; Let C be the estimated network loss power of the distribution network in time period t in the upper-level model; let C be the set of reactive power compensation devices, and Q be the estimated network loss power of the distribution network in time period t. c,t P represents the reactive power generated by reactive power compensation device c during time period t; d,t Q d,t Let d be the active and reactive power demand of the load during time period t.

[0217] (2) Lower-level model

[0218] The constraints of the lower-level model include cloud energy storage user constraints and distribution network power flow constraints. Cloud energy storage user constraints include electric vehicle constraints, uninterruptible power supply constraints, and demand response load constraints. Electric vehicle-related constraints mainly include electric vehicle charging and discharging capacity constraints, electric vehicle charging and discharging power constraints, electric vehicle minimum charging and discharging time constraints, and electric vehicle charging station total charging and discharging power limit constraints.

[0219]

[0220]

[0221]

[0222]

[0223]

[0224]

[0225]

[0226]

[0227]

[0228]

[0229]

[0230]

[0231]

[0232] In the formula: V represents the set of electric vehicles; These represent the charging and discharging power of electric vehicle v during time period t; O v,t Let η be the capacity of electric vehicle v during time period t. v Improve the charging and discharging efficiency of electric vehicles; These represent the upper and lower limits of the capacity of electric vehicle v, respectively. This refers to the initial battery level of electric vehicle V when it is connected to the grid. The expected charge level of electric vehicle v when it leaves the charging station; The net charge amount for electric vehicle v; These represent the upper and lower limits of the discharge power of the electric vehicle, respectively. Let v be the discharge state of electric vehicle v during time period t, where 1 indicates discharge and 0 indicates no discharge; These represent the upper and lower limits of the charging power of electric vehicles, respectively. The charging status of electric vehicle v during time period t is represented by 1, indicating charging and 0, indicating no charging. A counter for the offline time of electric vehicles after charging is complete. Minimum off-grid time after electric vehicle is fully charged; A counter for the off-grid time of an electric vehicle after it has been fully discharged. The minimum off-grid time for an electric vehicle after it has completely discharged its power; A counter for the on-grid charging time of electric vehicles. Minimum charging time for electric vehicles on the grid; For electric vehicles' on-grid discharge time counter, Let A(y) be the minimum discharge time of an electric vehicle on the grid; let A(y) be the set of equipment within electric vehicle charging station y, and let Y be the set of electric vehicle charging stations. These represent the maximum charging and discharging power of electric vehicle charging station y, respectively.

[0233] Demand response loads include interruptible loads, and the constraints are interruptible load limit constraints.

[0234]

[0235] In the formula: Let d be the interruptible load amount during time period t. α represents the participation status of interruptible load d during time period t; a value of 1 indicates it is in an interrupted state, and a value of 0 indicates it is not interrupted. d This refers to the proportion of interruptible load capacity to total load.

[0236] Uninterruptible power supply (UPS) related constraints include UPS capacity constraints and UPS charging and discharging power constraints.

[0237]

[0238]

[0239] O u,T =O u,0

[0240]

[0241]

[0242]

[0243] In the formula: U is the set of uninterruptible power supplies; These represent the charging and discharging power of the uninterruptible power supply u during time period t; O u,t Let η be the capacity of the uninterruptible power supply u during time period t. u For uninterruptible power supply charging and discharging efficiency; The upper and lower limits of the uninterruptible power supply (UPS) capacity; O u,T and O u,0 These represent the capacities of the uninterruptible power supply u at the end and beginning of the scheduling cycle, respectively. These represent the upper and lower limits of the discharge power of the uninterruptible power supply (UPS). The value represents the discharge state of the uninterruptible power supply u during time period t. A value of 1 indicates discharge, and a value of 0 indicates no discharge. These represent the upper and lower limits of the charging power of the uninterruptible power supply (UPS). The charging state of the uninterruptible power supply u during time period t is represented by a value of 1, indicating charging and a value of 0, indicating no charging.

[0244] The sum of the charging and discharging power of each cloud energy storage user in the lower-level model at each time period must be equal to the charging and discharging power of the cloud energy storage system at each time period obtained from the solution of the upper-level model.

[0245]

[0246]

[0247]

[0248] In the formula: A(e) is the set of devices in cloud energy storage system e.

[0249] A second-order cone relaxation method is used to linearize the power flow constraints of the distribution network. The operational constraints of the distribution network include power balance constraints at distribution network nodes, voltage drop limits on distribution lines, second-order cone constraints, and line operation constraints. Among these, the difference in network loss between the upper and lower level models is considered. Equal to the network loss value predicted by the upper-level model Compared with the actual network loss value of the lower-level model The difference.

[0250]

[0251]

[0252]

[0253]

[0254]

[0255]

[0256]

[0257]

[0258]

[0259] In the formula: A(i) is the set of distribution network equipment directly connected to node i; L, N bus Let be the set of transmission lines and nodes in the distribution network, respectively; s(l) and r(l) are the sending-end node and receiving-end node of transmission line l; P l,t Q l,t These represent the active and reactive power flow of transmission line l during time period t; The active and reactive power losses of transmission line l during time period t; ω represents the difference in network loss between the upper and lower layer models. l,t Let ψ be the square of the current in transmission line l during time period t. i,t r is the squared value of the voltage at node i during time period t; l and x l These are the resistance and reactance values ​​of transmission line l, respectively. These represent the squares of the upper and lower limits of the allowable current to flow through transmission line l, respectively. These represent the squares of the upper and lower limits of the voltage that node i can withstand, respectively.

[0260] Step 3: Establish an uncertainty model for the two-layer optimal scheduling of the power distribution network-cloud energy storage system, taking into account the uncertainties of wind and solar power.

[0261] When considering the uncertainty of wind and solar power output, typical wind and solar power output scenarios are incorporated into the upper-level model. Combined with the uncertainty model, a scheduling strategy that takes into account the uncertainty of wind and solar power output is obtained, and the scheduling strategy ensures that the distribution network can still operate stably when wind and solar power output fluctuates.

[0262] Example 3 presents an uncertainty model for establishing a two-layer optimal scheduling of a distribution network-cloud energy storage system that takes into account the uncertainties of wind and solar power:

[0263] Uncertain power output scenarios were generated using the Monte Carlo simulation method, and then the scenarios were reduced using the synchronous back-substitution method to obtain a set of typical wind and solar generator power output scenarios γ, with the scenario number sce.

[0264]

[0265]

[0266]

[0267]

[0268]

[0269]

[0270]

[0271]

[0272]

[0273]

[0274]

[0275]

[0276]

[0277]

[0278]

[0279]

[0280]

[0281]

[0282] In the formula: the superscript sce indicates the value of the variable under scenario sce, and γ represents the set of typical wind and solar power output scenarios; For the emergency ramp rate of the gas turbine unit; This refers to the adjustable range of substation transmission power under uncertain scenarios.

[0283] Step 4: Under the premise of satisfying the safety constraints of each component of the system, establish a two-layer optimization scheduling model of the distribution network-cloud energy storage system that takes into account the uncertainties of wind and solar power. The objective function of the upper-layer model is to minimize the operating cost of the distribution network. The optimal output plan of each power source in the distribution network and the charging and discharging plan of the cloud energy storage system are obtained by solving the model. The objective function of the lower-layer model is to maximize the profit of the cloud energy storage operator. Based on the charging and discharging plan of the cloud energy storage system obtained from the upper-layer model, and combined with the operating constraints of the distribution network, the charging and discharging plans of each electric vehicle user, uninterruptible power supply user, and demand response load user are formulated.

[0284] Example 4 presents the objective function of a two-layer optimal scheduling model for a distribution network-cloud energy storage system that considers the uncertainties of wind and solar power. In this model, the objective function of the upper-layer model is to minimize the operating cost of the distribution network, including the cost of using the cloud energy storage system, the cost of purchasing electricity from the upper-level grid, the cost of generating electricity from gas turbine units, and the penalty for renewable energy curtailment. The objective function expression is as follows:

[0285] minF1=F M +F E +F G +F Voll

[0286] In the formula: F1 is the operating cost of the distribution network, F M F represents the cost of electricity purchased by the distribution network from the upstream power grid. E For the cost of using cloud energy storage systems, F G For the cost of generating electricity from gas turbine units, F Voll The cost of curtailing renewable energy.

[0287] in:

[0288]

[0289] In the formula: Let t be the electricity purchase price corresponding to time period t.

[0290] When the power distribution network needs the cloud energy storage system to charge / discharge, it needs to pay the corresponding service fee to the cloud energy storage operator. The cloud energy storage operator only needs to pay the net charging fee to the power distribution network for the net charging amount of the cloud energy storage system.

[0291]

[0292] In the formula: The unit price for charging / discharging services of cloud energy storage system e. This represents the net charging unit price of the cloud energy storage system e.

[0293]

[0294] Where: M g (·) represents the heat rate curve. Let t be the fuel price per unit during time period t.

[0295]

[0296] In the formula: These are the unit prices for the curtailment penalties for photovoltaic generator sets and wind turbine generator sets, respectively.

[0297] In the lower-level model, the main component is the cloud energy storage system, and the objective function is to maximize the profit of the cloud energy storage operator. The profit of the cloud energy storage operator includes the following components:

[0298] max(F2-F P ) = F E -(F U +F V +F I )-F P

[0299] In the formula: F2 represents the total profit obtained by the cloud energy storage operator, which is equal to the cloud energy storage system usage cost paid by the distribution network to the cloud energy storage operator minus the cloud energy storage user dispatch cost paid by the cloud energy storage operator to the cloud energy storage user. U For uninterruptible power supply (UPS) user dispatch costs, F V For electric vehicle user dispatch costs, F I For interruptible load user scheduling costs, F P The penalty cost is the difference in network loss between the upper and lower layer models.

[0300] Cloud energy storage operators pay electric vehicle users for discharge / charge services, and electric vehicle users only need to pay for the net charging portion of the electric vehicle's net charging capacity.

[0301]

[0302] In the formula: The unit price for charging / discharging services for electric vehicles; Net charging cost per unit for electric vehicles.

[0303] Demand response load only considers interruptible loads of users. The cloud energy storage system uses a compensation method to induce users to reduce electricity consumption as part of the discharge power of the cloud energy storage system.

[0304]

[0305] In the formula: Provide compensation per unit price for interruptible loads to cloud energy storage system operators.

[0306] Cloud energy storage operators pay uninterruptible power supply (UPS) users for discharge / charge services.

[0307]

[0308] In the formula: The unit price for charging / discharging services for uninterruptible power supplies.

[0309] When the upper-level model solves for the power output plans of each unit in the distribution network and the charging and discharging plans of the cloud energy storage system, it does not calculate the power flow of the distribution network. Therefore, the network loss in the active power balance constraint is an estimated value. When the lower-level model formulates the charging and discharging plans of each cloud energy storage user, it performs distribution network power flow calculations to obtain the actual network loss value. The network loss difference is the difference between the actual network loss and the estimated network loss.

[0310]

[0311] In the formula: c p The penalty unit price is the difference in network loss.

[0312] Step 5: Use Monte Carlo simulation and synchronous back substitution to generate output uncertainty scenarios for wind and solar generator sets.

[0313] In the two-layer optimization scheduling model of the distribution network-cloud energy storage system considering the uncertainty of wind and solar power, the output plans of each unit obtained by the solution are based on the predicted output of wind and solar power. However, in reality, the output of wind and solar power is always accompanied by fluctuations, so its uncertainty needs to be considered. First, a large number of output uncertainty scenarios of wind and solar generator units are generated by Monte Carlo simulation method. Then, the scenarios are reduced by synchronous back substitution method. Finally, typical wind and solar power output scenarios are generated.

[0314] Example 5 provides a method for generating scenarios with uncertain power output from wind and solar generators:

[0315] Based on the basic prediction scenarios of wind and solar power generation, and assuming the wind power output prediction error... And photovoltaic power output prediction error It conforms to the standard normal distribution, that is Uncertainty-based power output scenarios were generated using the Monte Carlo simulation method. In scenario sce, the power outputs of wind power and solar power are as follows:

[0316]

[0317]

[0318] In the formula: and These represent the predicted power output values ​​of wind turbine w and photovoltaic generator s during time period t, respectively, under scenario sce. and These are the predicted output values ​​of wind turbine w and photovoltaic generator s in time period t, respectively, under the basic prediction scenario; and These represent the output prediction errors of wind turbine w and photovoltaic generator s during time period t, respectively, under scenario sce.

[0319] Because the large number of scenarios slows down the solution speed, a synchronous back-substitution method is used to reduce the number of scenarios, selecting a few typical scenarios for model solving to improve the solution efficiency. The steps for scenario reduction are as follows:

[0320] (1) Set SS as the initial scene set; DS is the scene set to be deleted, initially empty; Ω h (h = 1, 2, ..., N) represents the h-th output scenario in the initial scenario set SS, and scenario Ω h The corresponding probability of occurrence is ρ h Calculate the Euclidean distance DT between all scenes in the initial scene set SS. r,k :

[0321]

[0322] In the formula: DT r,k For the r-th scene Ω r and the kth scene Ω k The Euclidean distance between them, P t r and P t k The predicted power output of wind and solar turbines during time period t under scenarios r and k, respectively; Ω u This is a collection of all scenarios.

[0323] (2) PD k (r) represents scene Ω r The probabilistic distance to all other scenes is calculated. A smaller probabilistic distance indicates a higher similarity between the scene and other scenes. The probabilistic distance to each scene is calculated, and if the PD (probabilistic distance) is smaller... k (b) = minPD k If (r), (r=1,2,...,N), then choose scene Ω. b This is the scenario that will be deleted. The probability distance to PD is... k (r):

[0324]

[0325] (3) Calculate the relationship between each scene in the scene set SS and scene Ω. b The Euclidean distance, if DT b,d =minDT b,k (k≠b, b=1,2,...,N). Then delete scene Ω. b , will scene Ω b The probability of occurrence is added to scenario Ω d Up. Update the scene set and the probability of each scene appearing:

[0326] SS=SS-{Ω b};DS=DS+{Ω b}

[0327] ρ d =ρ d +ρ b ;ρ b =0

[0328] (4) Repeat (2) to (3) until the number of scenes meets the target setting value.

[0329] Step 6: Input the operating parameters of the distribution network-cloud energy storage system, and use a commercial solver to solve the two-layer optimal scheduling model of the distribution network-cloud energy storage system that takes into account the uncertainties of wind and solar power, so as to obtain its optimal scheduling strategy.

[0330] The operating parameters of the distribution network-cloud energy storage system in step 6 include the cost of output of each unit in the distribution network, ramp-up capability, minimum start-up and shutdown time, electrical load size, power transmission limit of transmission lines, current limit of transmission lines, voltage limit of nodes, charging and discharging service fee of cloud energy storage system, capacity limit of cloud energy storage system, charging and discharging power limit of cloud energy storage system, charging and discharging service fee of electric vehicle users, capacity limit of electric vehicle users, charging and discharging power limit of electric vehicle users, charging and discharging service fee of uninterruptible power supply (UPS) users, capacity limit of UPS users, charging and discharging power limit of UPS users, discharge compensation fee of demand response load, capacity limit of demand response load, etc.

[0331] The effects of the present invention will be described in detail below through specific embodiments.

[0332] (1) Example introduction.

[0333] like Figure 1 As shown, the distribution network system contains 33 nodes and 32 transmission lines; solid lines represent transmission circuits, and solid dots represent nodes; the numbers next to the solid dots are the node numbers; the connection of each unit is shown in the figure. The dual-layer optimization scheduling of the distribution network-cloud energy storage system, which takes into account the uncertainties of wind and solar power, studied in this example has a 24-hour research cycle and an interval of 1 hour.

[0334] (2) Analysis of the results of the examples.

[0335] To investigate the impact of different operating schemes of the distribution network-cloud energy storage system considering wind and solar power uncertainties on the system's economy and safety, the wind and solar power output uncertainty scenarios shown in Figures 2(a) and 2(b) were first extracted. The following two examples were then set up for comparative analysis: Example 1: Distribution network optimization scheduling including a cloud energy storage system; Example 2: Based on Example 1, considering the uncertainty of wind and solar power output.

[0336] The output of each generator unit in Example 1 is shown in Figures 3(a) and 3(b). In Example 1, the cloud energy storage system participates in the optimized scheduling of the distribution network. The cloud energy storage system mainly focuses on charging during off-peak hours (01:00–05:00 and 24:00) and discharging during peak hours (12:00–14:00 and 18:00–20:00). The distribution network load includes the base load and the cloud energy storage system charging load, with a peak-to-valley difference of 2.7 pu. In the upper-level model, the total operating cost of the distribution network is 93,401.2 yuan; of which the cost of purchasing electricity from the upper-level grid is 46,542.6 yuan; the cost of using the gas turbine unit is 45,382.1 yuan; the charging / discharging service fee paid to the cloud energy storage operator is 3,187.5 yuan, and the net charging fee is 1,711.0 yuan. In the lower-level model, the cloud energy storage operator pays electric vehicle users 399.1 yuan for charging / discharging services, collecting a net charging fee of 2037.0 yuan; the UPS charging / discharging service fee is 611.6 yuan; the load interruption compensation fee is 562.2 yuan; finally, the cloud energy storage operator's total revenue is 1940.6 yuan. Furthermore, electric vehicle users can also obtain certain benefits by participating in the cloud energy storage operation mode, significantly reducing their net charging costs.

[0337] The output of each generator set in Example 2 is as follows: Figure 4 As shown in Example 2, although the charging plan of the cloud energy storage system fluctuates significantly compared to the basic scenario under the uncertainty scenario, the overall trend remains unchanged. This is because, under the uncertainty scenario, the charging power of the cloud energy storage system at each time period is still constrained by the number of electric vehicles connected, the dispatchable UPS capacity, and the interruptible load capacity during that period. Under the influence of fluctuations in wind and solar power output, the charging and discharging of the cloud energy storage system have undergone significant changes. In Scenario 3, the wind turbine output increases between 21:00 and 23:00, and the cloud energy storage system absorbs excess wind power by increasing charging. In Scenario 5, the wind turbine output decreases at 02:00. At this time, the cloud energy storage system reduces its charging power and discharges to meet the load demand of the distribution network. The charging power of the cloud energy storage system at this time decreases from 0.6 pu in the basic scenario to 0.2 pu, and the discharging power increases from 0 p.u. in the basic scenario to 0.5 pu. This demonstrates that, while meeting the charging needs of cloud energy storage users, cloud energy storage systems can adjust the charging and discharging power at different times to cope with fluctuations in wind and solar power output, thereby improving the flexibility and stability of system operation.

[0338] Analysis of Examples 1 and 2 reveals that cloud energy storage systems can guide distributed energy storage resources on the user side to charge during off-peak hours and discharge during peak hours, reducing the operating costs of the distribution network and the net charging costs for cloud energy storage users, minimizing the peak-valley load difference, and achieving a win-win situation for the distribution network, cloud energy storage operators, and cloud energy storage users. The proposed dual-layer optimized operation model of the distribution network-cloud energy storage system, considering the uncertainties of wind and solar power output, incorporates uncertainty scenarios to address the uncertainties in wind and solar power output. By adjusting the charging and discharging of the cloud energy storage system, it ensures stable system operation during fluctuations in wind and solar power output.

[0339] The above description is merely a specific embodiment of the present invention, but it does not limit the scope of patent protection of the present invention. Any equivalent changes or substitutions made using the content of the present invention specification and drawings, and any direct or indirect application to other related technical fields, should be included within the scope of protection of the present invention.

Claims

1. A method for optimizing the scheduling of a distribution network-cloud energy storage system considering the uncertainties of wind and solar power, characterized in that, Includes the following steps: Step 1: Identify the various components of the distribution network-cloud energy storage system, including the distribution network, the cloud energy storage system, and cloud energy storage users. Cloud energy storage users include electric vehicle users, uninterruptible power supply users, and demand response load users. Step 2: Establish a deterministic model for the two-layer optimal scheduling of the distribution network-cloud energy storage system, taking into account the uncertainties of wind and solar power. The deterministic model includes an upper-level model and a lower-level model. The constraints of the upper-level model include constraints on cloud energy storage systems, constraints on distribution network units, and constraints on power balance. The constraints of the lower-level model include constraints on cloud energy storage users and constraints on distribution network power flow. The constraints on cloud energy storage users include constraints on electric vehicles, constraints on uninterruptible power supplies, and constraints on demand response loads. Step 3: Establish an uncertainty model for the two-layer optimal scheduling of the distribution network-cloud energy storage system, taking into account the uncertainties of wind and solar power. When considering the uncertainty of wind and solar power output, typical wind and solar power output scenarios are incorporated into the upper-level model. Combined with the uncertainty model, a scheduling strategy that considers the uncertainty of wind and solar power output is obtained, and the scheduling strategy ensures that the distribution network can still operate stably when wind and solar power output fluctuates. Step 4: Under the premise of satisfying the safety constraints of each component of the system, establish a two-layer optimization scheduling model of the distribution network-cloud energy storage system that takes into account the uncertainties of wind and solar power. The objective function of the upper-layer model is to minimize the operating cost of the distribution network, and the optimal output plan of each power source in the distribution network and the charging and discharging plan of the cloud energy storage system are obtained by solving the model. The objective function of the lower-layer model is to maximize the profit of the cloud energy storage operator. Based on the charging and discharging plan of the cloud energy storage system obtained by the upper-layer model, and combined with the operating constraints of the distribution network, the charging and discharging plans of each electric vehicle user, uninterruptible power supply user, and demand response load user are formulated. Step 5: Use Monte Carlo simulation to generate uncertain power output scenarios for wind and solar generators, then use synchronous back-substitution method to reduce the scenarios, and finally generate typical wind and solar power output scenarios. Step 6: Input the operating parameters of the distribution network-cloud energy storage system, and use a commercial solver to solve the two-layer optimal scheduling model of the distribution network-cloud energy storage system that takes into account the uncertainties of wind and solar power, so as to obtain its optimal scheduling strategy.

2. The optimized scheduling method for a distribution network-cloud energy storage system considering the uncertainties of wind and solar power, as described in claim 1, is characterized in that... In step 1, the power distribution network includes substation units, gas turbine units, wind and solar generator units, cloud energy storage systems, cloud energy storage users, electrical loads, bus nodes, and transmission lines.

3. The optimized scheduling method for a distribution network-cloud energy storage system considering the uncertainties of wind and solar power, as described in claim 1, is characterized in that... The constraints of the cloud energy storage system, distribution network unit, and power balance in the upper-level model in step 2 are specifically as follows: 1) Constraints of cloud energy storage systems The constraints of the cloud energy storage system include the charging and discharging power and capacity constraints of the cloud energy storage system. The cloud energy storage operator determines the upper and lower limits of the charging and discharging power of the cloud energy storage system in each time period and the upper and lower limits of the net charging power on the day by estimating the capacity limit of the cloud energy storage users under its management that can be dispatched in each time period. In the formula: t is the time index, T is the total scheduling period; e is the index of the cloud energy storage system, and E is the set of cloud energy storage systems in the distribution network; and Let t represent the discharge power and charging power of the cloud energy storage system e during time period t. and These represent the upper limits of discharge power and charging power of cloud energy storage system e during time period t, respectively. and These are the lower limits of discharge power and charging power for cloud energy storage system e during time period t, respectively. The net charge of cloud energy storage system e. and These represent the upper and lower limits of the net charging capacity of the cloud energy storage system; O e,t Let t be the capacity of cloud energy storage system e during time period t. and These are the upper and lower limits of the capacity of the cloud energy storage system, respectively. 2) Constraints of distribution network units The constraints on the power distribution network units include the operating constraints of gas turbine units, the output constraints of wind and solar turbine units, and the transmission power constraints of substations. a) Operating constraints of gas turbine units The operational constraints of gas turbine units specifically include upper and lower limits for output, minimum start-up time constraints, minimum shutdown time constraints, start-up and shutdown fuel consumption constraints, and ramp-up constraints; as shown in the following formulas: K g,t ≥k g ·(x g,t -x g,t-1 ),K g,t ≥0,g∈G,t∈T D g,t ≥d g ·(x g,t-1 -x g,t ),D g,t ≥0,g∈G,t∈T In the formula: g is the index of the gas turbine unit, G is the set of gas turbine units in the power distribution network; x g,t P represents the operating state of the gas turbine unit g during time period t, where 1 indicates it is in the start-up state and 0 indicates it is in the stop-down state; g,t and Q g,t These represent the active power and reactive power output by the gas turbine unit g during time period t, respectively. and These are the maximum and minimum limits for the active power output of the gas turbine unit, respectively. and These are the maximum and minimum limits for the reactive power output of the gas turbine unit, respectively. and These are the start-up and shutdown time counters for gas turbine unit g during time period t; and These represent the minimum start-up time and minimum shutdown time for the gas turbine unit, respectively; K g,t and D g,t The fuel consumption of the gas turbine unit during time period t for starting and stopping (g and k) are respectively. g and d g These represent the fuel consumption for the gas turbine unit g during startup and shutdown, respectively. and These are the uphill and downhill ramp rates for gas turbine unit g, respectively. b) Output constraints of wind and solar turbine units The output constraints of wind power and photovoltaic units include active power output constraints and reactive power output constraints; In the formula: S and W are the sets of photovoltaic generators and wind turbine generators in the distribution network, respectively. and P represents the predicted power output of photovoltaic generator s and wind turbine w during time period t. s,t and P w,t Q represents the actual output values ​​of photovoltaic generator s and wind turbine w during time period t. s,t and Q w,t Let t represent the reactive power output of photovoltaic generator s and wind turbine generator w during time period t. and The power factor angles of photovoltaic generator s and wind turbine generator w are respectively. c) Substation transmission power constraints Substation transmission power constraints include active power limit constraints and reactive power limit constraints; In the formula: Z represents the set of substations in the distribution network. The power purchased by the distribution network from the upper-level power grid through substation z during time period t. The reactive power supplied by the upstream power grid to the distribution network through substation z during time period t. and These represent the upper and lower limits of active power transmitted from the upstream power grid to the distribution network through substation z, respectively. and These are the upper and lower limits of reactive power transmitted from the upper-level power grid to the distribution network through substation z, respectively. 3) Power balance constraints Power balance constraints include active power balance constraints and reactive power balance constraints: In the formula, D represents the set of nodes where the load is located; P t loss Let C be the estimated network loss power of the distribution network in time period t in the upper-level model; let C be the set of reactive power compensation devices, and Q be the estimated network loss power of the distribution network in time period t. c,t P represents the reactive power generated by reactive power compensation device c during time period t; d,t and Q d,t Let d be the active power demand and reactive power demand of load d during time period t.

4. The optimized scheduling method for a distribution network-cloud energy storage system considering the uncertainties of wind and solar power as described in claim 1, characterized in that, The cloud energy storage user constraints and distribution network power flow constraints in the lower-level model in step 2 are specifically as follows: 1) Cloud energy storage user constraints Cloud energy storage user constraints include electric vehicle constraints, demand response load constraints, and uninterruptible power supply constraints. a) Electric vehicle constraints Electric vehicle constraints include electric vehicle charging and discharging capacity constraints, electric vehicle charging and discharging power constraints, electric vehicle minimum charging and discharging time constraints, and electric vehicle charging station total charging and discharging power limit constraints. In the formula: V is the set of electric vehicles; and These represent the charging power and discharging power of electric vehicle v during time period t; O v,t Let η be the capacity of electric vehicle v during time period t. v Improve the charging and discharging efficiency of electric vehicles; and These represent the upper and lower limits of the capacity of electric vehicle v, respectively. This refers to the initial battery level of electric vehicle V when it is connected to the grid. The expected charge level of electric vehicle v when it leaves the charging station; The net charge amount for electric vehicle v; and These represent the upper and lower limits of the discharge power (v) of the electric vehicle, respectively. Let v be the discharge state of electric vehicle v during time period t, where 1 indicates discharge and 0 indicates no discharge; and These represent the upper and lower limits of the charging power of electric vehicle V, respectively. The charging status of electric vehicle v during time period t is represented by 1, indicating charging and 0, indicating no charging. A counter for the offline time of an electric vehicle after it has been fully charged during time period t. Minimum off-grid time after electric vehicle is fully charged; This is a counter for the duration of time an electric vehicle remains off-grid after it has fully discharged during time period t. The minimum off-grid time for an electric vehicle after it has completely discharged its power; This is a counter for the on-grid charging time of electric vehicles during period t. Minimum charging time for electric vehicles on the grid; This is a counter for the on-grid discharge duration of electric vehicles during time period t. Let A(y) be the minimum discharge time of an electric vehicle on the grid; let A(y) be the set of equipment within electric vehicle charging station y, and let Y be the set of electric vehicle charging stations. and These represent the maximum charging power and discharging power of electric vehicle charging station y, respectively. b) Demand response load constraints Demand response loads include interruptible loads, and the constraint is the interruptible load limit constraint. In the formula: Let d be the interruptible load amount during time period t. α represents the participation status of interruptible load d during time period t; a value of 1 indicates it is in an interrupted state, and a value of 0 indicates it is not interrupted. d This refers to the proportion of interruptible load capacity to total load. c) Uninterruptible power supply constraints Uninterruptible power supply (UPS) constraints include UPS capacity constraints and UPS charging and discharging power constraints. The u,T =O u,0 In the formula: U is the set of uninterruptible power supplies; and These represent the charging power and subsequent discharging power of the uninterruptible power supply u during time period t; O u,t Let η be the capacity of the uninterruptible power supply u during time period t. u For uninterruptible power supply charging and discharging efficiency; and The upper and lower limits of the uninterruptible power supply capacity (U); O u,T and O u,0 These represent the capacities of the uninterruptible power supply u at the end and beginning of the scheduling cycle, respectively. and These represent the upper and lower limits of the discharge power of the uninterruptible power supply (UPS), respectively. This represents the discharge state of the uninterruptible power supply u during time period t. A value of 1 indicates discharge, and a value of 0 indicates no discharge. and These represent the upper and lower limits of the charging power of the uninterruptible power supply (UPS), respectively. The charging state of the uninterruptible power supply u during time period t is represented by 1, which indicates charging and 0 indicates no charging. The sum of the charging and discharging power of each cloud energy storage user in the lower-level model at each time period must be equal to the charging and discharging power of the cloud energy storage system at each time period obtained from the upper-level model. In the formula: A(e) is the set of devices in cloud energy storage system e; 2) Distribution network power flow constraints A second-order cone relaxation method is used to linearize the power flow constraints of the distribution network. The operational constraints of the distribution network include power balance constraints at distribution network nodes, voltage drop limits on distribution lines, second-order cone constraints, and line operation constraints. Among these, the difference in network losses between the upper and lower level models is considered. Equal to the network loss value P predicted by the upper-level model t loss Compared with the actual network loss value of the lower-level model The difference; In the formula: A(i) is the set of distribution network equipment directly connected to node i; L and N bus Let be the set of transmission lines and nodes in the distribution network; s(l) and r(l) are the sending-end node and receiving-end node of transmission line l; P l,t and Q l,t These represent the active power flow and reactive power flow of transmission line l during time period t, respectively. and The active and reactive power losses of transmission line l during time period t; This represents the difference in network loss between the upper and lower layer models. ω l,t Let ψ be the square of the current in transmission line l during time period t. j,t and ψ i,t Let r be the squared voltage values ​​of node j and node i respectively during time period t; l and x l These are the resistance and reactance values ​​of transmission line l, respectively. and ψ represents the squares of the upper and lower limits of the allowable current flowing through transmission line l, respectively. i max and ψ i min These represent the squares of the upper and lower voltage limits that node i can withstand, respectively.

5. The optimized scheduling method for a distribution network-cloud energy storage system considering the uncertainties of wind and solar power as described in claim 1, characterized in that, In step 3, the uncertainty model for wind and solar power output is as follows: In the formula: the superscript sce indicates the value of the variable under scenario sce, and γ represents the set of typical wind and solar power output scenarios; For the emergency ramp rate of the gas turbine unit; This refers to the adjustable range of substation transmission power under uncertain scenarios.

6. The optimized scheduling method for a distribution network-cloud energy storage system considering the uncertainties of wind and solar power as described in claim 1, characterized in that, In step 4 of the distribution network-cloud energy storage system dual-layer optimization scheduling model (1) In the upper-level model, the distribution network operation cost in the objective function includes the cost of using the cloud energy storage system, the cost of purchasing electricity from the upper-level grid, the cost of generating electricity from gas turbine units, and the penalty for renewable energy curtailment. The objective function expression is as follows: minF1=F M +F E +F G +F Voll In the formula: F1 is the operating cost of the distribution network, F M F represents the cost of electricity purchased by the distribution network from the upstream power grid. E For the cost of using cloud energy storage systems, F G For the cost of power generation by gas turbine units, F Voll The cost of curtailing renewable energy; in: In the formula: The electricity purchase price corresponding to time period t; When the distribution network needs the cloud energy storage system to charge / discharge, it pays the corresponding service fee to the cloud energy storage operator, while the cloud energy storage operator only needs to pay the net charging fee to the distribution network for the net charging amount of the cloud energy storage system. In the formula: The unit price for charging / discharging services of cloud energy storage system e. The net charging unit price for cloud energy storage system e; Where: M g (·) represents the heat rate curve. Let be the fuel unit price during time period t; In the formula: and These are the unit prices for the curtailment penalties for photovoltaic power generation units and wind power generation units, respectively. (2) In the lower-level model, the main body is the cloud energy storage system, and the profit of the cloud energy storage operator in the objective function includes the following parts: max(F2-F P )=F E -(F U +F V +F I )-F P In the formula: F2 represents the total profit obtained by the cloud energy storage operator, which is equal to the cloud energy storage system usage cost paid by the distribution network to the cloud energy storage operator minus the cloud energy storage user dispatch cost paid by the cloud energy storage operator to the cloud energy storage user. U For uninterruptible power supply (UPS) user dispatch costs, F V For electric vehicle user dispatch costs, F I For interruptible load user scheduling costs, F P The penalty cost is the difference in network loss between upper and lower layer models. Cloud energy storage operators pay electric vehicle users for discharge / charge services, and electric vehicle users only need to pay for the net charging portion of the electric vehicle's net charging capacity. In the formula: The unit price for charging / discharging services for electric vehicles; Net charging cost per unit for electric vehicles; Demand response load only considers the interruptible load of users. The cloud energy storage system uses a compensation method to induce users to reduce electricity consumption as part of the discharge power of the cloud energy storage system. In the formula: Provide cloud energy storage system operators with a compensation price for interruptible loads; Cloud energy storage operators pay uninterruptible power supply (UPS) users for discharge / charging services; In the formula: Unit price for charging / discharging services for uninterruptible power supplies; When the upper-level model solves the power output plan of each unit in the distribution network and the charging and discharging plan of the cloud energy storage system, it does not calculate the power flow of the distribution network. Therefore, the network loss in the active power balance constraint is an estimated value. When the lower-level model formulates the charging and discharging plan of each cloud energy storage user, it performs power flow calculation of the distribution network to obtain the actual network loss value. The network loss difference is the difference between the actual network loss and the estimated network loss. In the formula: c p The penalty unit price is the difference in network loss.

7. The optimized scheduling method for a distribution network-cloud energy storage system considering the uncertainties of wind and solar power as described in claim 1, characterized in that, The specific process of step 5 is as follows: Based on the basic prediction scenarios of wind and solar power generation, and assuming the wind power output prediction error... And photovoltaic power output prediction error It conforms to a standard normal distribution, that is Uncertainty-based power output scenarios were generated using the Monte Carlo simulation method. In scenario sce, the power outputs of wind power and solar power are as follows: In the formula: and These represent the predicted power output values ​​of wind turbine w and photovoltaic generator s during time period t, respectively, under scenario sce. and These are the predicted output values ​​of wind turbine w and photovoltaic generator s in time period t, respectively, under the basic prediction scenario; and These are the output prediction errors of wind turbine w and photovoltaic generator s during time period t, respectively, under scenario sce; The scene reduction method is used to select several typical scenes for model solving in order to improve the solution efficiency. The scene reduction steps are as follows: Step a: Set SS as the initial scene set; DS is the scene set to be deleted, initially empty; calculate the Euclidean distance between each scene in the initial scene set SS: In the formula: DT r,k For the r-th scene Ω r and the kth scene Ω k The Euclidean distance between them, P t r and P t k The predicted power output of wind and solar turbines during time period t under scenarios r and k, respectively; Ω u A collection of all scenarios; Step b: Calculate the scenario Ω r Probability distance (PD) from all other scenarios k (r), the smaller the value, the higher the similarity between the scene and other scenes; In the formula: ρ r For scene Ω r The probability of occurrence; r = 1, 2, ..., N, where N is the total number of scenes; Step c: Calculate the probability distance between each scene. If there exists a scene Ω... b Probability distance (PD) from all other scenarios k (b) = minPD k (r); then select scene Ω b , as a scenario that will be deleted; Step d: Calculate the relationship between each scene and scene Ω in the initial scene set SS. b The Euclidean distance, if DT b,d =minDT b,k If k ≠ b, and b = 1, 2, ..., N, then delete scene Ω. b , will scene Ω b The probability of occurrence is added to scenario Ω d Up; Update the scene set and the probability of each scene occurring: SS=SS-{Ω b };DS=DS+{Ω b } r d =ρ d +r b ;r b =0 Step e: Repeat steps bd until the number of scenes meets the target set value.

8. The optimized scheduling method for a distribution network-cloud energy storage system considering the uncertainties of wind and solar power as described in claim 1, characterized in that, The operating parameters of the distribution network-cloud energy storage system include the cost of output of each unit in the distribution network, ramp-up capability, minimum start-up and shutdown time, electrical load size, power transmission limit of transmission lines, current limit of transmission lines, voltage limit of nodes, charging and discharging service fee of cloud energy storage system, capacity limit of cloud energy storage system, charging and discharging power limit of cloud energy storage system, charging and discharging service fee of electric vehicle users, capacity limit of electric vehicle users, charging and discharging power limit of electric vehicle users, charging and discharging service fee of uninterruptible power supply (UPS) users, capacity limit of UPS users, charging and discharging power limit of UPS users, discharge compensation fee of demand response load, and capacity limit of demand response load.

Images