Distributed energy aggregation flexibility analysis method suitable for distribution network and micro-grid levels

Abstract

The invention discloses a distributed resource aggregation flexibility analysis method suitable for a distribution network and a micro-grid level. The method comprises the following steps: acquiring scene data and distributed resource data of distributed photovoltaic and load; constructing a scene uncertainty set according to the scene data, and determining operation constraints of distributed resources according to the distributed resource data; based on distributed resource operation constraints and a scene uncertainty set, constructing a deterministic optimization problem with flexibility maximization of the micro-grid nodes and the distribution network nodes as a target, and based on a deterministic optimization result and an uncertainty set, constructing a double-layer two-stage robust optimization model with flexibility maximization of the micro-grid nodes and the distribution network nodes as a target; and solving the double-layer two-stage robust optimization model to obtain the power flexibility adjustment range of the micro-grid node and the distribution network node. According to the method, the uncertainty of a historical scene of distributed resource operation and the uncertainty of a superior scheduling instruction are considered, and double-layer two-stage robust optimization with the flexibility maximization of micro-grid and distribution network nodes as a target is constructed on the basis, so that the robustness of the system can be improved on the premise of ensuring the robustness of the system. And the flexibility range of the distributed resources is expanded and the time sequence distribution is optimized.

Description

Technical Field

[0001] The present invention relates to the aggregation flexibility analysis technology of distributed resources, belonging to the technical field of robust dispatching optimization of power grids with high-proportion distributed energy access, and particularly relates to a method for analyzing the aggregation flexibility of distributed resources applicable to the distribution network and microgrid levels. Background Art

[0002] As the traditional source-load structure is gradually replaced by the new source-load structure of distribution-microgrid coordination, the traditional deterministic optimization method can no longer evaluate and utilize the improved flexibility brought about by the transformation of the roles of distribution networks and microgrids from traditional energy consumers to energy producers and consumers, resulting in a waste of the flexible regulation ability of distributed resources and further increasing the operating cost. Therefore, according to the regulation characteristics of distributed resources and by using time-series scenarios to quantify the typical operating conditions of loads and photovoltaics in microgrids, the flexible regulation ability of distribution networks and microgrids can be effectively characterized. At the same time, the traditional two-stage robust optimization model should be effectively improved to reduce the overly conservative tendency of traditional robust optimization while ensuring system stability and make full use of the flexibility of distributed resources in the system. Summary of the Invention

[0003] The purpose of the present invention is to provide a method for analyzing the aggregation flexibility of distributed resources applicable to the distribution network and microgrid levels. By combining the mechanism modeling of distributed resources with historical power information and cooperating with the robust optimization method, the flexible regulation ability of distributed resources at the distribution network and microgrid levels is analyzed, and the flexibility range of distributed resources can be expanded and its time-series distribution can be optimized while ensuring the robustness of the system. The technical solution adopted by the present invention is as follows.

[0004] On the one hand, the present invention provides a method for analyzing the aggregation flexibility of distributed resources applicable to the distribution network and microgrid levels, including:

[0005] Obtaining the scenario data of distributed photovoltaics and loads in the region, as well as the distributed resource data;

[0006] Constructing a scenario uncertainty set according to the scenario data, and determining the operation constraints of distributed resources according to the distributed resource data;

[0007] Based on the operation constraints of the distributed resources and the scenario uncertainty set, constructing a deterministic optimization problem with the goal of maximizing the flexibility of microgrid nodes and distribution network connection points, and solving to obtain a deterministic optimization result;

[0008] Based on the deterministic optimization result and the scenario uncertainty set, constructing a two-layer two-stage robust optimization model with the goal of maximizing the flexibility of microgrid nodes and distribution network connection points;

[0009] Solve the double - layer two - stage robust optimization model to obtain the node power flexibility intervals of the distribution network and the micro - grid.

[0010] For the day - ahead scheduling problem, through the present invention, the time - series data of the node power flexibility adjustment intervals of the distribution network and the micro - grid within 24 hours of the next scheduling period can be obtained, providing a reference for actual scheduling. It can not only accurately evaluate the flexible adjustment capabilities of distributed resources at the distribution network and micro - grid levels, expand the flexibility range and optimize its time - series distribution, but also ensure the system robustness.

[0011] Optionally, the distributed resource data includes but is not limited to the number of distributed resources, types, and the adjustable range of a single distributed resource;

[0012] The types of distributed resources include micro - gas turbines, energy storage, and adjustable loads, and the models are respectively established as follows:

[0013] (1) Micro - gas turbine:

[0014] (1a)

[0015] (1b)

[0016] (1c)

[0017] (1d)

[0018] In the formula, is the output of the gas turbine at node i at time t, and are respectively the upper and lower limits of the output of the gas turbine at node i; and are the up - ramp and down - ramp values of the gas turbine at node i at time t, and are the maximum limits of the up - ramp and down - ramp of the gas turbine at node i at time t, represents a single standard time interval;

[0019] (2) Energy storage:

[0020] (2a)

[0021] (2b)

[0022] (2c)

[0023] (2d)

[0024] In the formula, and is the charging and discharging power of the energy storage at node i at time t, and are the charging efficiency and discharging efficiency of the energy storage respectively; is the upper limit of the charging and discharging power of the energy storage at node i, is the upper limit of the energy storage capacity;

[0025] (3) Adjustable load:

[0026] (3a)

[0027] (3b)

[0028] In the formula, is the upward adjustment amount of the flexible load aggregation power at the i-th node in the t-th period, and are the maximum upward adjustment amount and downward adjustment amount of the flexible load aggregation power at the i-th node respectively; is the initial power consumption of the flexible load at the i-th node, is the minimum power consumption of the flexible load.

[0029] The above modeling is the basis for the subsequent deterministic optimization and two-layer two-stage robust optimization of the present invention.

[0030] The present invention is mainly applicable to the day-ahead scheduling of the power grid. Therefore, optionally, the scenario data of the distributed photovoltaic and the load include the historical operation data of the distributed photovoltaic and the load, the historical meteorological data, and the meteorological prediction data;

[0031] Construct a scenario uncertainty set according to the scenario data, including:

[0032] For the distributed photovoltaic and the load at each node in the analysis area, analyze the power probability distribution characteristics respectively according to the historical operation data, the historical meteorological data, and the meteorological prediction data, obtain the power probability density fitting functions of the distributed photovoltaic and the load, and generate scenarios based on the sampling method to obtain the upper and lower boundary modeling of the scenario uncertainty set;

[0033] Among them, the scenario uncertainty set of the distributed photovoltaic and the load is expressed as:

[0034] (4)

[0035] Among them, n indicates that the description object of the uncertainty set is the distributed photovoltaic or the load ; represents the scenario uncertainty set; Z is a set used to control the occurrence of uncertain extreme situations, and the elements in it and They are the upper and lower boundaries of the worst-case scenario, both of which are 0-1 variables; is the set uncertainty adjustment parameter, representing the maximum number of operating intervals when uncertainty encounters the worst-case scenario within a scheduling period. For day-ahead scheduling, it is usually ; 、 represent the true value and predicted value of the uncertain variable respectively, is the maximum deviation of the fluctuation range of the historical or generated scenario of distributed PV or load, expressed as:

[0036] (5)

[0037] In the formula, and are the scenario sets of distributed PV and load respectively, 、 are the power value and average value of the node distributed PV or load at time t in the s-th scenario respectively. The generation of the scenario uncertainty set can adopt existing technologies. Among them, the historical scenario can be obtained from historical data, and the generated scenario for the future period is obtained according to historical data and predicted data for the future period, which is not the content concerned in this application and will not be elaborated.

[0038] Optionally, the deterministic optimization problem aiming at maximizing the flexibility of microgrid nodes and distribution network connection points includes a deterministic optimization model for maximizing the flexibility of microgrid port nodes and a deterministic optimization model for maximizing the flexibility of distribution network connection point nodes, where:

[0039] The deterministic optimization model for maximizing the flexibility of microgrid port nodes is expressed as:

[0040] (6a)

[0041] (6b)

[0042] (6c)

[0043] (6d)

[0044] The deterministic optimization model for maximizing the flexibility of distribution network connection point nodes is expressed as:

[0045]

[0046] (7a)

[0047] ​​​ , (7c)

[0048] , (7d)

[0049] , (7e)

[0050] (7f)

[0051] (7g)

[0052] (7h)

[0053] (7i)

[0054] Among them, constraints (6b) and (7b) are the operating boundary constraints of distributed resources, representing the operating constraints of micro gas turbines, energy storage, and adjustable loads in distributed resources, which are (1a)-(1d), (2a)-(2d), (3a)-(3b) introduced previously; constraints (7c)-(7h) are the linearized constraints of the distribution network power flow; is the flexibility reward for the flexible regulation interval of the node at time t;

[0055] i and j are any adjacent nodes in the distribution network except the gateway nodes, is the set of energy storage nodes, is the set of PV nodes, is the set of load nodes; , are the maximum and minimum active power values of the gateway node i of the microgrid, respectively; , are the maximum and minimum active power values of the distribution network gateway node 1, respectively; is the reactive power value of node i; and are the upper and lower limits of the charge and discharge power of the energy storage at node i; and are the upper and lower limits of the output of distributed PV at node i; and are the upper and lower limits of the load at node i; , are the voltage amplitude and phase angle of distribution network node i, respectively; , are the power value and average value of distributed PV at node i, respectively; , They are respectively the power value and the average value of the load on node i.

[0056] Optionally, the linearized constraint of the distribution network power flow is obtained from the relationship between the voltage amplitude phase angle and the injected power of the distribution network nodes as follows:

[0057] (8a)

[0058] (8b)

[0059] where N is the number of distribution network nodes, represents the voltage amplitudes of N distribution network nodes, is the voltage phase angle of N distribution network nodes, and are respectively the injected active power and the injected reactive power of N distribution network nodes;

[0060] For the given IEEE 33-node system, node 1 is used as the balancing node for the distribution network power flow calculation, and the intermediate parameters are as follows: , , the real part of the admittance matrix of node 1 , the real part and the imaginary part of the admittance matrix of node 2 、 , the real part and the imaginary part of the admittance matrix of node 3 、 , the real part and the imaginary part of the admittance matrix of node 4 、 . The elements on the diagonal of the admittance matrix are the self-admittances of the nodes corresponding to the subscripts, and the remaining elements are the mutual admittances between the nodes corresponding to the subscripts.

[0061] Based on the relationships in the above (8a) and (8b), the optimization solution based on the constraints (7c)-(7h) can be realized, so as to ensure that the reactive power, voltage amplitude, and phase angle of each node in the distribution network are within a reasonable range.

[0062] For the above deterministic optimization model, existing solution methods such as the Gurobi solver can be used for solution. After the deterministic optimization solution, the benchmark intervals for the power flexibility regulation of the microgrid and distribution network nodes and the corresponding optimal regulation incentives can be obtained under the condition of ignoring the uncertainties of distributed photovoltaic and load. The benchmark intervals for the node power flexibility regulation can be expressed as:

[0063]

[0064] where is the set of microgrid port nodes. The benchmark intervals for the node power flexibility regulation and the corresponding optimal regulation incentives obtained by the deterministic optimization will be used as the constraints for the subsequent two-layer two-stage robust optimization.

[0065] The deterministic optimization result under the premise of ignoring the uncertainties of photovoltaic and load in the present invention is expressed as .

[0066] Optionally, the two-layer two-stage robust optimization model aiming at maximizing the flexibility of microgrid nodes and distribution network connection points includes a microgrid two-layer two-stage robust optimization model and a distribution network two-layer two-stage robust optimization model, where:

[0067] The microgrid two-layer two-stage robust optimization model is expressed as:

[0068] (9a)

[0069] (9b)

[0070] (9c)

[0071] (9d)

[0072] (9e)

[0073] (9f)

[0074] The distribution network two-layer two-stage robust optimization model is expressed as:

[0075] (10a)

[0076] (10b)

[0077] (10c)

[0078] (10d)

[0079] (10e)

[0080] (10f)

[0081] (10g)

[0082] (10h)

[0083] In the formula, represents the robustness level of the uncertain parameter, is the objective deviation factor, is the set of uncertain scenarios of the upper-level power dispatch instruction, is the scenario element in ; is the optimal value of the flexibility regulation incentive obtained by deterministic optimization solution, is the optimization objective of the lower-level model, which represents the flexibility regulation incentive of uncertainty optimization, and are the adjustable load and rigid load at time t on node i respectively;

[0084] In the microgrid two-layer two-stage robust optimization model, (9a)-(9b) is the upper-level model for determining the maximum robust levels of PV and load, and (9c)-(9f) is the lower-level model for determining the node power regulation range that can obtain the maximum incentive in the flexible interval of the microgrid. In the lower-level model, (9c) is the objective function for the two-stage optimization of the microgrid, (9d) is the first-stage constraint, and (9e)-(9f) are the second-stage constraints;

[0085] In the distribution network two-layer two-stage robust optimization model, (10a)-(10b) is the upper-level model, and (10c)-(10g) is the lower-level model. In the lower-level model, (10c) is the objective function for the two-stage optimization of the distribution network, (10d) is the first-stage constraint, and (10e)-(10g) are the second-stage constraints.

[0086] The above-mentioned set of uncertain scenarios of the upper-level power dispatch instruction can be a random set of upper-level dispatch instructions.

[0087] When the present invention performs the regulation flexibility interval planning for the microgrid and the distribution network based on the above two-layer two-stage robust optimization model, the optimization of the power flexibility interval of the distribution network needs to be realized under the condition that the flexibility interval of the microgrid is known.

[0088] Optionally, the solution of the two-layer two-stage robust optimization model includes:

[0089] For the upper-level models in the microgrid two-layer two-stage robust optimization model and the distribution network two-layer two-stage robust optimization model, regarding them as problems involving the solution of a unary equation system and using the random bisection method to obtain the value of the system robust coefficient ;

[0090] For the lower-level models in the microgrid two-layer two-stage robust optimization model and the distribution network two-layer two-stage robust optimization model, based on the system robust coefficient obtained by solving the upper-level model , the C&CG algorithm is used to iteratively decompose the original optimization problem into a master problem and a sub-problem, obtain the optimal solution of the original problem, and obtain the time-series distribution data of the power flexibility interval corresponding to the maximized flexibility interval reward.

[0091] Optionally, the solution using the random bisection method includes:

[0092] Step S11: Given the objective deviation factor and the uncertainty adjustment parameter , use the Gurobi solver to solve the deterministic optimization model to obtain the optimal value of the flexibility regulation incentive ;

[0093] Step S12: Select a random number in the interval , set the robustness level , where , ; Substitute into the lower-layer two-stage robust optimization model, and use the C&CG algorithm and the Gurobi solver to solve the lower-layer model to obtain the optimal solution of the lower-layer model, denoted as ;

[0094] Step S13: If , then set , otherwise set , and go to Step S14;

[0095] Step S14: If , stop the iteration, return the robustness level and the optimal solution , otherwise return to Step S12.

[0096] In a second aspect, the present invention provides a computer-readable storage medium, on which a computer program is stored. When the computer program is executed by a processor, the steps of the distributed resource aggregation flexibility analysis method applicable to the distribution network and microgrid levels as introduced in the first aspect are implemented.

[0097] Beneficial effects

[0098] The present invention takes into account the scenario uncertainty of using node distributed resources and the uncertainty of superior dispatching instructions. On this basis, a flexible and adjustable two-layer two-stage robust optimization model with the maximization of the flexibility of microgrid nodes and distribution network gateways as the goal is constructed. It can expand the flexibility range of distributed resources and optimize its time-series distribution while ensuring the robustness of the system, so as to effectively analyze or predict the flexibility interval of the regulation of microgrids and distribution networks with high penetration of distributed resources. Description of the drawings

[0099] Figure 1 The figure shows a schematic diagram of the collaborative framework for flexibility optimization at the distribution network and microgrid levels;

[0100] Figure 2 The figure shows a schematic diagram of the IEEE 33-node system;

[0101] Figure 3 The figure shows a schematic diagram of the process of an embodiment of the method of the present invention;

[0102] Figure 4 The figure shows a schematic diagram of the process of solving the upper-layer model using the random bisection method. Detailed implementation manners

[0103] The following is a further description in conjunction with the accompanying drawings and specific embodiments.

[0104] The collaborative framework for flexibility optimization at the distribution network and microgrid levels is as Figure 1 shown. The microgrid accesses the distribution network through the microgrid gateway node, and the distribution network receives the superior dispatching instructions through the distribution network gateway node. Among them, distributed resources, loads, and superior dispatching instructions all have uncertainties. Therefore, the present invention aims to consider these uncertainty factors, and perform a two-layer two-stage robust optimization with the maximization of the flexibility of microgrid nodes and distribution network gateways as the goal, effectively analyze or predict the flexibility intervals of the regulation of microgrids and distribution networks with a high penetration rate of distributed resources, and expand the flexibility range of distributed resources and optimize their temporal distribution on the premise of ensuring the robustness of the system.

[0105] Embodiment 1

[0106] This embodiment introduces a method for analyzing the aggregation flexibility of distributed resources applicable to the distribution network and microgrid levels. Referring to Figure 3 , the method includes:

[0107] Obtain the scenario data of distributed photovoltaics and loads in the region, as well as the distributed resource data;

[0108] Construct a scenario uncertainty set according to the scenario data, and determine the operation constraints of distributed resources according to the distributed resource data;

[0109] Based on the operation constraints of the distributed resources and the scenario uncertainty set, construct a deterministic optimization problem with the maximization of the flexibility of microgrid nodes and distribution network gateways as the goal, and solve to obtain the deterministic optimization result;

[0110] Based on the deterministic optimization result and the scenario uncertainty set, construct a two-layer two-stage robust optimization model with the maximization of the flexibility of microgrid nodes and distribution network gateways as the goal;

[0111] Solve the two-layer two-stage robust optimization model to obtain the node power flexibility intervals of the distribution network and microgrid.

[0112] The specific implementation of this embodiment includes the following content.

[0113] I. Mechanism Modeling of Distributed Resources

[0114] The distributed resource data required to be obtained in this embodiment includes, but is not limited to, the number and types of distributed resources within the research area and the adjustable range of individual distributed resource units;

[0115] The types of distributed resources include micro gas turbines, energy storage, and adjustable loads, and the modeling is as follows:

[0116] (1) Micro gas turbine:

[0117] (1a)

[0118] (1b)

[0119] (1c)

[0120] (1d)

[0121] In the formula, is the output of the gas turbine at node i at time t, and are the upper and lower limits of the output of the gas turbine at node i, respectively; and are the up-ramp and down-ramp values of the gas turbine at node i at time t, and are the maximum limits of the up-ramp and down-ramp of the gas turbine at node i at time t, represents a single standard time interval;

[0122] (2) Energy storage:

[0123] (2a)

[0124] (2b)

[0125] (2c)

[0126] (2d)

[0127] In the formula, and are the charge and discharge powers of the energy storage at node i at time t, 、 are the energy storage charging efficiency and energy storage discharging efficiency, respectively; is the upper limit value of the charge and discharge power of the energy storage at node i, is the upper limit value of the energy storage capacity;

[0128] (3) Adjustable load:

[0129] (3a)

[0130] (3b)

[0131] Wherein, is the upward adjustment amount of the flexible load aggregation power of the i-th node at the t-th time period, and are respectively the maximum upward adjustment amount and the downward adjustment amount of the flexible load aggregation power of the i-th node; is the initial power consumption of the flexible load of the i-th node, is the minimum power consumption of the flexible load.

[0132] The above modeling is the basis for the subsequent deterministic optimization and the two-layer two-stage robust optimization of the present invention.

[0133] II. Construction of Scenario Uncertainty Set

[0134] In order to implement the day-ahead scheduling of the power grid, the scenario data of distributed photovoltaic and load required in this embodiment includes the historical operation data of distributed photovoltaic and load, the historical meteorological data, and the meteorological prediction data.

[0135] Construct a scenario uncertainty set according to the scenario data, including:

[0136] For the distributed photovoltaic and load of each node in the analysis area, respectively analyze the power probability distribution characteristics according to the historical operation data, the historical meteorological data, and the meteorological prediction data, obtain the power probability density fitting function of distributed photovoltaic and load, and generate scenarios based on the sampling method to obtain the upper and lower boundary modeling of the scenario uncertainty set;

[0137] Among them, the scenario uncertainty sets of distributed photovoltaic and load are expressed as:

[0138] (4)

[0139] Where n represents that the description object of the uncertainty set is distributed photovoltaic or load ; represents the scenario uncertainty set; Z is a set used to control the occurrence of uncertain extreme situations, and the elements therein , are respectively the upper and lower boundaries of the worst-case scenario, and are both 0-1 variables; is a set uncertainty adjustment parameter, representing the maximum number of operating intervals for the worst-case scenario of uncertainty within a scheduling period. For day-ahead scheduling, it is usually ; 、 represent the true value and predicted value of the uncertain variable respectively, is the maximum deviation of the fluctuation range of the historical or generated scenario of distributed PV or load, expressed as:

[0140] (5)

[0141] In the formula, and are the scenario sets of distributed PV and load respectively, 、 are the power sampling value and average value of node distributed PV or load respectively. The generation of the scenario uncertainty set can adopt existing technologies. Among them, the historical scenario can be obtained from historical data, and the generated scenario for future periods is obtained based on historical data and predicted data for future periods.

[0142] III. Modeling of the distribution and microgrid collaborative system

[0143] Referring to Figure 2 the IEEE 33-node system shown, "MG", "PV", and "ES" represent the microgrid access node, PV node, and energy storage node respectively.

[0144] Ignoring the shunt admittance of each node, denoting the real part of the node admittance matrix as and the imaginary part as , the matrix form of the linearized power flow equation can be written as:

[0145] (13a)

[0146] Node 1 is the slack node, , , we get:

[0147] (13b)

[0148] (13c)

[0149] In the formula, and are the block matrices of and respectively: , , , ; , , , 。

[0150] Equation (8) can be derived from Equation (13c):

[0151] (8a)

[0152] (8b)

[0153] where , 。

[0154] Meanwhile, several other types of special nodes in the distribution network are defined:

[0155] 1) Energy storage node set : The active power injected into the node is equal to the discharge power of the ES, satisfying Equations (2a) and (2d), and serving as a controllable variable for distribution network-level optimization;

[0156] 2) Photovoltaic node set and load node set : According to the fluctuation ranges of the load and photovoltaic output, the uncertainty set of the node active power is as shown in Equation (4);

[0157] 3) Microgrid connection node set : Since the microgrid effectively integrates the distributed resources within its control range through its energy management system, the node-level flexibility range presented externally is:

[0158] (8c)

[0159] where is the decision variable for the microgrid-level two-layer two-stage flexibility interval optimization.

[0160] The distribution network power flow linearization constraint can be derived from Equation (8).

[0161] IV. Construction of the deterministic optimization model

[0162] In this embodiment, it is first assumed that the predicted values of the distributed photovoltaic and load are the same as the actual values, and a deterministic optimization model for maximizing the node flexibility range of the microgrid port i and the distribution network gateway node l is constructed and solved:

[0163] (6a)

[0164] (6b)

[0165] (6c)

[0166] (6d)

[0167] The deterministic optimization model for maximizing the flexibility of distribution network gateway nodes is expressed as:

[0168] (7a)

[0169] (7b)

[0170] , (7c)

[0171] , (7d)

[0172] , (7e)

[0173] (7f)

[0174] (7g)

[0175] (7h)

[0176] (7i)

[0177] Among them, constraints (6b) and (7b) are the operation boundary constraints of distributed resources, representing the operation constraints of micro gas turbines, energy storage, and adjustable loads in distributed resources; constraints (7c)-(7h) are the linearization constraints of distribution network power flow; is the flexibility reward for the flexible adjustment interval of nodes at time t;

[0178] i and j are any adjacent nodes in the distribution network except the gateway nodes, is the set of energy storage nodes, is the set of photovoltaic nodes, is the set of load nodes, is the set of microgrid connection nodes; , are the maximum and minimum active power values of the microgrid gateway node respectively; , are the maximum and minimum active power values of the distribution network gateway node respectively; is the reactive power value of node i; is the upper limit value of the charging and discharging power of the energy storage on node i; is the lower limit value of the output of distributed resources on node i; is the lower limit of the load on node i; and are the voltage magnitude and phase angle of distribution network node i, respectively; and are the power value and average value of distributed PV on node i, respectively; and are the power value and average value of the load on node i, respectively.

[0179] For the above deterministic optimization problem, given the objective deviation factor and the uncertainty adjustment parameter , using the Gurobi solver to solve the deterministic optimization model, the optimal value of the flexibility regulation incentive of the maximum time series flexibility interval of the microgrid can be obtained on the premise of ignoring the uncertainties of PV and load .

[0180] V. Construction and solution of the two-layer two-stage robust optimization model

[0181] Establish a method for maximizing the flexibility interval based on two-stage robust optimization to address the uncertainties of distributed PV and load power. Its objective is to maximize the volatility of uncertain parameters while meeting the predetermined objectives. The robustness of the model can be controlled by adjusting the parameter . The mathematical model of the flexible two-stage robust optimization is as follows:

[0182] (9a)

[0183] (9b)

[0184] (9c)

[0185] (9d)

[0186] (9e)

[0187] (9f)

[0188] Similarly, establish a two-stage robust optimization model with flexible flexibility of the power at the distribution network gateway:

[0189] (10a)

[0190] (10b)

[0191] (10c)

[0192] (10d)

[0193] (10e)

[0194] (10f)

[0195] (10g)

[0196] (10h)

[0197] Taking the microgrid flexibility model as an example, in the upper-layer models (9a)-(9b), are the objective values obtained from deterministic optimization respectively, is the objective deviation factor, which is used to represent the predetermined objective value for the expansion of the flexibility interval when considering the uncertainties of load and photovoltaic output and the objective value obtained from deterministic optimization the degree of deviation between them. represents the robustness level of the uncertain parameters. In the lower-layer models (9c)-(9f), is the uncertain set of the superior power dispatch instruction, and there is . Equation (9c) is the objective function for the two-stage optimization of the microgrid, equation (9d) is the first-stage constraint, and the others are the second-stage constraints. and are the first-stage control variables, is the unit reward for the flexible regulation interval of the node at different time periods, which can affect the time-series distribution of the regulation ability and reflects the demand of the superior dispatcher for flexibility at different time periods. The larger it is, the greater the contribution of flexibility to the overall optimization of the superior dispatch at this time period. There is no optimization objective in the second stage, and its purpose is to ensure the effective execution of the dispatch instruction after the uncertainty is revealed. The executability is provided by constraint (9e).

[0198] In the uncertain set U, is subjectively set by the decision maker, which represents the maximum number of operating intervals for the worst-case scenario of uncertainty in one cycle. Since the uncertainties and prediction deviations of wind power and photovoltaic power generation are independent, the probability that a single uncertain parameter reaches its boundary simultaneously is very low, usually . By solving the adjustable two-stage robust model with a fixed value and the objective deviation factor, It represents the maximum load and PV power deviation that the system can tolerate when the worst-case scenario occurs during the maximum Γ operating interval within the operating cycle. Compared with traditional stochastic optimization, the adjustable two-stage robust optimization uses a polyhedral uncertainty set to model uncertain parameters without the need to know the probability distribution function of the uncertain parameters, thus achieving higher computational efficiency. In addition, compared with traditional robust optimization, the adjustable two-stage robust optimization provides greater flexibility and allows adjustment of objective factors and the uncertainty adjustment parameter to control the conservatism of the model. Therefore, the adjustable two-stage robust optimization method proposed in the present invention can quantitatively analyze the relationship between the range of variation of uncertain parameters and the predicted flexibility regulation objectives of the distribution network and microgrid under the condition of limited budget regulation cost.

[0199] For the adjustable two-stage robust optimization model in this embodiment, a modified coupled CCG algorithm of the stochastic bisection method is constructed for solution. Based on the lower-layer two-stage robust model of the flexibility interval expansion problem of the flexible adjustable two-stage robust optimization, the optimal planning solution under the worst-case scenario u is solved by the C&CG algorithm. The C&CG algorithm iteratively decomposes the original problem into a master problem and a sub-problem to obtain the optimal solution of the original problem; the upper-layer model of the flexibility interval expansion problem based on the flexible adjustable two-stage robust optimization is a problem involving the solution of a unary equation system and is solved using the stochastic bisection method. Specifically:

[0200] For the upper-layer models in the microgrid double-layer two-stage robust optimization model and the distribution network double-layer two-stage robust optimization model, they are regarded as problems involving the solution of a unary equation system and the system robust coefficient is obtained by solving using the stochastic bisection method;

[0201] For the lower-layer models in the microgrid double-layer two-stage robust optimization model and the distribution network double-layer two-stage robust optimization model, based on the system robust coefficient obtained by solving the upper-layer model, the C&CG algorithm is used to iteratively decompose the original optimization problem into a master problem and a sub-problem to obtain the optimal solution of the original problem, and the time-series distribution data of the power flexibility interval corresponding to the maximized flexibility interval reward is obtained.

[0202] The upper-layer model is solved using the stochastic bisection method as Figure 4 follows, specifically including:

[0203] Step S11: Given the objective deviation factor and the uncertainty adjustment parameter , the deterministic optimization model is solved using the Gurobi solver to obtain the optimal value of the flexibility regulation incentive;

[0204] Step S12: In the interval Select a random number from , and set the robustness level , where , ; Substitute into the lower-level two-stage RO model, and use the C&CG algorithm and the Gurobi solver to solve the lower-level model to obtain the optimal solution of the lower-level model, denoted as ;

[0205] Step S13: If , then set , otherwise set , and go to Step S14;

[0206] Step S14: If , stop the iteration, and return the robustness level and the optimal solution , otherwise return to Step S12.

[0207] Solve the lower-level model using the C&CG algorithm, including:

[0208] S21, write the two-level two-stage robust optimization model in a compact form:

[0209] (15a)

[0210] (15b)

[0211] (15c)

[0212] (15d)

[0213] (15e)

[0214] (15f)

[0215] (15g)

[0216] where, in the lower-level two-stage robust optimization model, is the first-stage decision variable, is the second-stage decision variable, is the first-stage objective function, is the second-stage objective function, is the uncertain variable, (15d) is the first-stage constraint, and (15e)-(15g) are the second-stage constraints, and is its dual variable and can be set artificially;

[0217] S22 divides the master - subproblem for the lower - layer two - stage robust optimization. The master problem is expressed as:

[0218] (16a)

[0219] (16b)

[0220] , (16c)

[0221] , (16d)

[0222] where \(l\) is the current iteration number, is the solution of the sub - problem obtained in the \(l\) - th iteration, is the worst - case scenario obtained in the \(l\) - th iteration;

[0223] The sub - problem is expressed as:

[0224] (17a)

[0225] (17b)

[0226] (17c)

[0227] (17d)

[0228] where, is the feasible region of the two - stage variable, determined by equations (1)-(3), is the optimal solution of the robust optimization master - problem decision;

[0229] S23, considering that the lower - layer two - stage robust sub - problem model is linear, according to the strong duality theory, the min - max problem is transformed into a single - layer min problem, expressed as:

[0230] (18a)

[0231] (18b)

[0232] Substitute (4) and (17g) into it, and linearize the bilinear term using the big - M method to obtain the result:

[0233] (19a)

[0234] , (19b)

[0235] , (19c)

[0236] , (19d)

[0237] , , (19e)

[0238] where M is a positive real number large enough, and are introduced auxiliary variables.

[0239] Relative to the lower-level robust optimization model, is a known quantity provided by the upper-level model. Through the above derivations and transformations, the master problem and the sub-problem become mixed-integer programming problems and can be effectively solved by the Gurobi solver.

[0240] So far, the nodal power flexibility intervals of the distribution network and the microgrid considering the uncertainties of distributed energy, load, and superior dispatching instructions, and their corresponding optimal regulation incentives can be obtained.

[0241] Embodiment 3

[0242] Based on the same inventive concept as Embodiment 1, this embodiment introduces a computer-readable storage medium on which a computer program is stored. When the computer program is executed by a processor, it implements the distributed resource aggregation flexibility analysis method applicable to the distribution network and microgrid levels as described in Embodiment 1.

[0243] Those skilled in the art should understand that the embodiments of the present application can be provided as a method, a system, or a computer program product. Therefore, the present application can take the form of a complete hardware embodiment, a complete software embodiment, or an embodiment combining software and hardware aspects. Moreover, the present application can take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0244] This application is described with reference to the flowcharts and / or block diagrams of methods, apparatuses (systems), and computer program products according to embodiments of the present application. It should be understood that each flow and / or block in the flowchart and / or block diagram, and the combination of flows and / or blocks in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to the processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing devices to generate a machine, so that the instructions executed by the processor of the computer or other programmable data processing devices generate a means for implementing the functions specified in one or more flows and / or blocks in the process Figure 1 one or more flows and / or blocks Figure 1 a means for implementing the functions specified in one or more blocks or multiple blocks

[0245] These computer program instructions can also be stored in a computer-readable memory that can direct a computer or other programmable data processing device to work in a specific manner, so that the instructions stored in the computer-readable memory generate a manufactured article including an instruction means, and the instruction means implements the functions specified in one or more flows and / or blocks in the process Figure 1 one or more flows and / or blocks Figure 1 a means for implementing the functions specified in one or more blocks or multiple blocks

[0246] These computer program instructions can also be loaded onto a computer or other programmable data processing device, so that a series of operation steps are executed on the computer or other programmable device to generate a computer-implemented process, and thus the instructions executed on the computer or other programmable device provide steps for implementing the functions specified in one or more flows and / or blocks in the process Figure 1 one or more flows and / or blocks Figure 1 a means for implementing the functions specified in one or more blocks or multiple blocks

[0247] The embodiments of the present invention have been described above in conjunction with the accompanying drawings. However, the present invention is not limited to the above specific embodiments. The above specific embodiments are merely illustrative and not restrictive. Under the inspiration of the present invention, those of ordinary skill in the art can also make many forms without departing from the spirit and scope protected by the present invention and the claims. These all belong to the protection scope of the present invention

Claims

1. A method for analyzing the flexibility of distributed resource aggregation applicable to the distribution network and microgrid levels, characterized in that Including: Obtain the scenario data of distributed photovoltaic and load within the area, as well as the distributed resource data; Construct a scenario uncertainty set according to the scenario data, and determine the operation constraints of distributed resources according to the distributed resource data; Based on the operation constraints of the distributed resources and the scenario uncertainty set, construct a deterministic optimization problem aiming at maximizing the flexibility of microgrid nodes and distribution network connection points, and solve to obtain a deterministic optimization result; Based on the deterministic optimization result and the scenario uncertainty set, construct a two-layer two-stage robust optimization model aiming at maximizing the flexibility of microgrid nodes and distribution network connection points; Solve the two-layer two-stage robust optimization model to obtain the node power flexibility intervals of the distribution network and the microgrid.

2. The method according to claim 1, characterized in that, The distributed resource data includes the quantity and type of distributed resources and the adjustable range of each distributed resource unit; The types of distributed resources include micro gas turbines, energy storage, and adjustable loads, and their models are as follows: (1) Micro gas turbine: （1a） （1b） （1c） （1d） In the formula, is the output of the gas turbine at the ith node at time t, and are the upper and lower limits of the output of the gas turbine at the ith node, respectively; and are the upward and downward ramp values of the gas turbine at the ith node at time t, and are the maximum limits of the upward and downward ramps of the gas turbine at the ith node at time t, represents a single standard time interval; (2) Energy storage: （2a） （2b） （2c） （2d） Wherein, and are the charging and discharging powers of the energy storage at node i at time t, , are the charging efficiency and discharging efficiency of the energy storage respectively; is the upper limit value of the charging and discharging power of the energy storage at node i, is the upper limit value of the energy storage capacity; (3) Adjustable load: （3a） （3b） In the formula, is the upward adjustment amount of the flexible load aggregated power of the i-th node in the t-th period, and are the maximum upward adjustment amount and the downward adjustment amount of the flexible load aggregated power of the i-th node respectively; is the initial power consumption of the flexible load of the i-th node, is the minimum power consumption of the flexible load.

3. The method according to claim 2, characterized in that, The scenario data of the distributed photovoltaic and load includes the historical operation data of the distributed photovoltaic and load, the historical meteorological data, and the meteorological prediction data; Construct a scenario uncertainty set according to the scenario data, including: For the distributed photovoltaic and load at each node within the analysis area, respectively analyze the probability distribution characteristics of node power according to the historical operation data, historical meteorological data, and meteorological prediction data, obtain the power probability density fitting functions of the distributed photovoltaic and load, and generate scenarios based on the sampling method to obtain the upper and lower boundary models of the scenario uncertainty set; Among them, the scenario uncertainty set of the distributed photovoltaic and load is expressed as: （11） where n represents that the description object of the uncertainty set is distributed photovoltaic or load ; represents the scenario uncertainty set; Z is a set used to control the occurrence of extreme uncertainty situations, and the elements in it , are the upper and lower boundaries of the worst-case scenario respectively, both of which are 0-1 variables; is the set uncertainty adjustment parameter, representing the maximum number of operating intervals when uncertainty encounters the worst-case scenario within a scheduling period; , represent the true value and predicted value of the uncertainty variable respectively, is the maximum deviation of the fluctuation range of the historical or generated scenario of distributed photovoltaic or load, expressed as: （12） In the formula, and are the scenario sets of distributed PV and load respectively, , are the power value and average value of the node distributed PV or load at the t-th moment in the s-th scenario respectively.

4. The method according to claim 3, wherein The deterministic optimization problem aiming at maximizing the flexibility of microgrid nodes and distribution network connection points includes a deterministic optimization model for maximizing the flexibility of microgrid port nodes and a deterministic optimization model for maximizing the flexibility of distribution network connection points, where: Microgrid port node The deterministic optimization model for maximizing flexibility is expressed as: （6a） （6b） （6c） （6d） Distribution network gateway node The deterministic optimization model for maximizing flexibility is expressed as: （7a） （7b） ， （7c） ， （7d） ， （7e） （7f） （7g） （7h） （7i） Among them, constraints (6b) and (7b) are the operating boundary constraints of distributed resources, representing the operating constraints of micro gas turbines, energy storage, and adjustable loads in distributed resources; constraints (7c)-(7h) are the linearization constraints of the distribution network power flow; is the flexibility reward for the flexible adjustment interval of nodes at time t. In practical applications, this flexibility reward can be taken as the electricity price; i and j are any adjacent nodes except the gateway nodes in the distribution network, is the set of energy storage nodes, is the set of photovoltaic nodes, is the set of load nodes; , are the maximum and minimum active power of the gateway node i of the microgrid, respectively; , are the maximum and minimum active power of the gateway node 1 of the distribution network, respectively; is the reactive power value of node i; and are the upper and lower limits of the energy storage charge and discharge power on node i; and are the upper and lower limits of the output of distributed photovoltaics on node i; and are the upper and lower limits of the load on node i; , are the voltage amplitude and phase angle of distribution network node i, respectively; , are the power value and average value of distributed photovoltaics on node i, respectively; , are the power value and average value of the load on node i, respectively.

5. The method according to claim 4, characterized in that The linearized constraint of the distribution network power flow is obtained from the relationship between the voltage amplitude phase angle of the distribution network node and the node injection power as follows: （8a） （8b） where N is the number of distribution network nodes, represents the voltage magnitudes of the N nodes in the distribution network, is the voltage phase angle of the N nodes in the distribution network, and are the injected active power and injected reactive power of the N nodes in the distribution network respectively; For the given IEEE 33-node system, node 1 is used as the slack node for distribution network power flow calculation, and the intermediate parameters are as follows: , , the real part of the admittance matrix of node 1 , the real and imaginary parts of the admittance matrix of node 2 、 , the real and imaginary parts of the admittance matrix of node 3 、 , the real and imaginary parts of the admittance matrix of node 4 、 .

6. The method according to claim 5, characterized in that, The two-layer two-stage robust optimization model aiming at maximizing the flexibility of microgrid nodes and distribution network connection points includes a microgrid two-layer two-stage robust optimization model and a distribution network two-layer two-stage robust optimization model, where: The microgrid two-layer two-stage robust optimization model is expressed as: (9a) (9b) (9c) (9d) (9e) (9f) The distribution network two-layer two-stage robust optimization model is expressed as: (10a) (10b) (10c) (10d) (10e) (10f) (10g) (10h) In the formula, represents the robustness level of uncertain parameters, is the objective deviation factor, is the set of uncertain scenarios of the upper-level power dispatch instruction, is the scenario element in ; is the optimal value of the flexibility regulation incentive obtained by deterministic optimization solution, is the optimization objective of the lower-level model, which represents the flexibility regulation incentive of uncertainty optimization, and are the adjustable load and the rigid load at node i at time t respectively; In the microgrid two-layer two-stage robust optimization model, (9a)-(9b) are the upper-layer model, (9c)-(9f) are the lower-layer model. In the lower-layer model, (9c) is the objective function for the two-stage optimization of the microgrid, (9d) is the first-stage constraint, and (9e)-(9f) are the second-stage constraints; In the distribution network two-layer two-stage robust optimization model, (10a)-(10b) are the upper-layer model, (10c)-(10g) are the lower-layer model. In the lower-layer model, (10c) is the objective function for the two-stage optimization of the microgrid, (10d) is the first-stage constraint, and (10e)-(10g) are the second-stage constraints.

7. The method according to claim 6, characterized in that, Solving the two-layer two-stage robust optimization model includes: For the upper-level models in the microgrid two-layer two-stage robust optimization model and the distribution network two-layer two-stage robust optimization model, regard them as problems involving the solution of a unary equation system and use the random dichotomy method to solve for the system robust coefficient value; For the lower-level models in the microgrid two-layer two-stage robust optimization model and the distribution network two-layer two-stage robust optimization model, based on the system robust coefficient obtained by solving the upper-level model , the C&CG algorithm is used to iteratively decompose the original optimization problem into a master problem and a subproblem, obtain the optimal solution of the original problem, and obtain the time-series distribution data of the power flexibility interval corresponding to the maximum flexibility interval reward.

8. The method according to claim 7, wherein The solution using the stochastic bisection method includes: Step S11: Given an objective deviation factor and an uncertainty adjustment parameter , use the Gurobi solver to solve the deterministic optimization model and obtain the optimal value of the flexibility adjustment incentive ; Step S12: Select a random number in the interval and set the robustness level , where ; Substitute into the lower-level two-stage RO model, and use the C&CG algorithm and the Gurobi solver to solve the lower-level model to obtain the optimal solution of the lower-level model, denoted as ; ; Step S13: If , then set , otherwise set , and go to Step S14; Step S14: If , stop the iteration and return the robustness level and the optimal solution , otherwise return to Step S12.

9. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by a processor, it implements the steps of the distributed resource aggregation flexibility analysis method according to any one of claims 1-8.

Images