Uncertainty optimization method for active distribution network considering reconfiguration and user participation

By constructing the current constraints of active distribution networks that include switch reconstruction and user participation, combined with the robust optimization scheduling model of multi-scenario distribution, the economic and safety problems of active distribution networks under wind and light uncertainty are solved, and the economy and safety of active distribution networks are achieved. Optimized operation of the distribution system.

CN117856246BActive Publication Date: 2024-11-05成都思拓知识产权服务有限公司
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202410030994.5
Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
Filing Date
2024-01-09
Publication Date
2024-11-05
Estimated Expiration
2044-01-09

AI Technical Summary

Technical Problem

When facing the uncertainty of scenery, the existing active distribution networks are difficult to fully utilize the demand response and adjustment capabilities under reconstruction and user participation, resulting in insufficient economic and safety of the distribution system operation.

Method used

By constructing the active distribution network trend constraints including switch reconstruction, combining the demand response constraints under different types of user participation, a multi-scenario distribution robust optimization scheduling model is built, and Latin hypercube sampling and improved K-means clustering is adopted. The algorithm generates typical scenarios, and uses column and constraint generation algorithms to solve them to optimize the distribution network operation.

Benefits of technology

It improves the economy and safety of the power distribution system, fully utilizes the demand response adjustment capabilities under switch reconstruction and user participation, optimizes the coordinated coordination of wind and light uncertainty, and reduces grid loss and voltage fluctuations.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN117856246B_ABST
    Figure CN117856246B_ABST
Patent Text Reader

Abstract

The present invention discloses an active distribution network uncertainty optimization method considering reconstruction and user participation, which belongs to the technical field of active distribution network optimization operation, and comprises the following steps: step S1: constructing an active distribution network flow constraint including switch reconstruction based on the active distribution network flow constraint condition; step S2: considering the demand response under different types of user participation based on limited publicity resources, establishing a functional relationship between different types of user participation and demand response publicity resources, and constructing a corrected demand response constraint; step S3: aiming at the uncertainty of wind and solar power, taking the penalty cost of zone voltage fluctuation and the operating cost as the objective function, constructing a multi-scenario distributed robust optimization scheduling model for the active distribution network; step S4: scenario generation and reduction, and solving the model with a column and constraint generation algorithm. The present invention can coordinate the switch reconstruction with the demand response under different types of user participation, and effectively improve the economy and safety of the operation of the distribution system.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present invention belongs to the technical field of active distribution network optimization operation, and in particular relates to an active distribution network uncertainty optimization method considering reconstruction and user participation. Background Art

[0002] In recent years, due to energy shortages and climate change, the scale of clean energy such as wind power and photovoltaic power grid connection has continued to increase, resulting in a series of problems such as local voltage increase, power flow reversal, and increased network losses in the distribution network system. In order to solve the above problems, active distribution network technology with flexibility and controllability has emerged. Active distribution network mainly coordinates and controls distributed power sources, energy storage devices, demand response loads and other resources, which can effectively improve the economy and safety of the distribution system.

[0003] In order to cope with the increasing risks and challenges of active distribution networks in the future, it is necessary to explore more flexible control resources to participate in the optimization operation of active distribution networks, among which switch reconstruction can be used as one of the important means of regulation. Existing research on switch reconstruction mainly focuses on reducing network losses and improving voltage levels, and there are few studies on the comprehensive consideration of economic, safe and reliable operation of distribution networks, which cannot fully play its benefits in the comprehensive operation of active distribution networks. Demand response, as an important regulation resource of active distribution networks, has the function of encouraging users to adjust their own electricity consumption behavior and improve the load curve. The degree of user's subjective participation in demand response regulation is closely related to the amount of demand response control resources, and different types of users have different participation in demand response regulation. There is still a lack of research on how to regulate the demand response level of various types of users under the premise of limited publicity resources, and the regulation capacity of demand response under different types of user participation cannot be fully utilized. In view of the uncertainty of wind and solar, both traditional robust optimization and stochastic optimization have limitations. The distributed robust optimization based on multiple scenarios has the advantages of both, and it has a good balance between economy and robustness. Under the premise of considering the uncertainty of wind and solar, with economy and safety as the goal, how to give full play to the advantages of reconstruction and demand response under different types of user participation and the potential of the two in coordination is of great significance. Therefore, there is a need for an active distribution network uncertainty optimization method that takes reconstruction and user participation into consideration and can effectively improve the economy and safety of distribution system operation. Summary of the invention

[0004] The purpose of the present invention is to provide an active distribution network uncertainty optimization method taking into account reconstruction and user participation, which can effectively improve the economic performance and safety of distribution system operation.

[0005] To achieve the above purpose, the technical solution adopted by the present invention is:

[0006] An active distribution network uncertainty optimization method considering reconstruction and user participation includes the following steps:

[0007] Step S1: constructing active distribution network power flow constraints including switch reconstruction based on active distribution network power flow constraints;

[0008] Step S2: Considering the demand response under different types of user participation based on limited publicity resources, establishing a functional relationship between different types of user participation and demand response publicity resources, and constructing a modified demand response constraint condition;

[0009] Step S3: In view of the uncertainty of wind and solar power, the penalty cost of voltage fluctuation in different zones and the operating cost are taken as the objective function, reactive support and energy storage are added, and a multi-scenario distributed robust optimization dispatch model of active distribution network is constructed according to the power flow constraints including switch reconstruction, demand response constraints under different types of user participation, radial structure constraints, safe operation constraints and regulation resource operation constraints;

[0010] Step S4: Use Latin hypercube sampling and improved K-means clustering algorithm to generate multiple typical scenarios, and use column and constraint generation algorithm to solve the model to obtain the active distribution network optimization scheduling plan.

[0011] A further improvement of the technical solution of the present invention is that in step S1, the active distribution network flow constraint condition including switch reconstruction is constructed based on the active distribution network flow constraint condition as follows:

[0012]

[0013] In the formula, r ij is the equivalent resistance of branch ij; x ij is the equivalent reactance of branch ij; P j,t and Q j,t are the active and reactive power injected by node j during period t; α ij,t is the on-off state variable of branch ij in period t, which is a Boolean variable. The variable value 0 indicates disconnection and 1 indicates closing. ij,t and Q ij,t are the active and reactive power flowing through branch ij during period t respectively; I ij,t is the current value of line ij during period t; V i,t is the voltage amplitude of node i at time period t; P ij,max , Q ij,max ,I ij,maxare respectively the square of the current amplitude of branch ij in period t, the square of the voltage amplitude of node i in period t, the maximum active power borne by branch ij, the maximum reactive power borne by branch ij, and the maximum square of the current amplitude borne by branch ij; f(j) is the set of branch head nodes with j as the end node; s(j) is the set of branch end nodes with j as the head node; ψ E Represents the set of all branches in the distribution network.

[0014] A further improvement of the technical solution of the present invention is that in step S2, the functional relationship between the participation of different types of users and the demand response publicity resources is:

[0015]

[0016]

[0017] In the formula, x d is the dead zone threshold; x f Saturation threshold; s i is user engagement; x i Promote resources for demand response; 1,dr , ψ 2,dr and ψ 3,dr are the load nodes for demand response in residential areas, commercial areas and industrial areas respectively; x n For fixed publicity resources, it is regarded as a constant.

[0018] When the publicity resources are lower than the minimum dead zone threshold, the user participation is 0. When the publicity resources are greater than the minimum dead zone threshold and less than the minimum saturation threshold, the user participation shows a linear growth trend. When the publicity resources are greater than the minimum saturation threshold, the user participation does not increase with the change of publicity resources and always maintains a value of 1.

[0019] A further improvement of the technical solution of the present invention is that: in step S2, the modified demand response constraint condition includes an interruptible load constraint and a shiftable load constraint, wherein:

[0020] Interruptible load constraints:

[0021]

[0022] In the formula, and are respectively the active and reactive interruption decision variables of the interruptible load at node i in period t; and are respectively the upper limits of active and reactive interruption of the interruptible load at node i in period t;

[0023] Translational load constraints:

[0024]

[0025]

[0026] In the formula, is the load level of the translatable load at node i during period t; is the total power demand of the shiftable load at node i during the scheduling period; and are the minimum load level and maximum load level of the load that can be translated by node i in period t.

[0027] A further improvement of the technical solution of the present invention is that in step S3, the radial structure constraint is:

[0028]

[0029] In the formula, β ij,t It represents the flow direction of the current in the ij branch during time period t. When the value is 1, it means that the current flows from j to i. H(i) is the set of all nodes adjacent to node i.

[0030] A further improvement of the technical solution of the present invention is that: in step S3, the safe operation constraints include node voltage constraints and branch current constraints, wherein:

[0031] Node voltage constraints:

[0032]

[0033] Branch current constraints:

[0034]

[0035] Where V i,min and V i,max are the minimum and maximum voltage allowed at node i respectively.

[0036] A further improvement of the technical solution of the present invention is that in step S3, the resource operation constraints include wind power and photovoltaic output constraints, switch opening and closing times constraints, energy storage system operation constraints, discrete reactive compensation device constraints, and continuous reactive compensation constraints, wherein:

[0037] Wind power and photovoltaic output constraints:

[0038]

[0039] In the formula, and are the wind and solar outputs of node i in period t respectively; and They are the corresponding wind power and photovoltaic predicted power respectively;

[0040] Switch opening and closing times constraints:

[0041]

[0042] In the formula, ψ E is the set of all branches; c max Indicates the total number of times the switch is opened and closed within the cycle; c d,max Indicates the total number of times a single switch is opened and closed within a cycle;

[0043] Energy storage system operation constraints:

[0044]

[0045]

[0046]

[0047] In the formula, and are the charging sign and discharging sign at node i at time t respectively; and are the charging power and discharging power at node i at time t respectively; and They are the upper limits of the active charging and discharging power of ESS respectively; is the ESS capacity at node i during period t; and Respectively represent the minimum and maximum values ​​of the ESS capacitance at node i; η loss , η ch and η dis They represent the electric energy self-consumption rate, charging efficiency and discharging efficiency respectively;

[0048] Discrete reactive power compensation device constraints:

[0049]

[0050]

[0051]

[0052] In the formula, ψ CB It is the collection of CB grid connection points; is the reactive compensation power of CB at node i at time t; is the number of CB grid-connected groups, which is a discrete variable; is the reactive compensation power of each CB group at node i at time t; and They are grid-connected action identifiers, which are Boolean variables; The maximum number of switching times within a cycle; is the maximum number of CB switching groups at each time at node i;

[0053] Constraints for continuous reactive power compensation:

[0054]

[0055] In the formula, ψ SVC It is a collection of SVC grid connection points; represents the reactive compensation power of SVC at node i at time t; and They represent the minimum and maximum values ​​of reactive power compensation of SVC at node i respectively.

[0056] A further improvement of the technical solution of the present invention is that in step S3, the objective function including the penalty cost of voltage fluctuation in the partition and the operating cost includes two-stage costs, the first-stage cost includes the cost of switch opening and closing and the cost of energy storage cycle, and the second-stage cost includes the cost of wind and solar abandonment, energy storage charging and discharging cost, network loss cost, demand response cost and partition voltage fluctuation cost, and the objective function expression is:

[0057]

[0058] In the formula, x is the first-stage variable; y s is the second-stage variable in the sth scenario; p s is the probability of each scenario; ψ=ψ1∩ψ ∞ is the set of scene probability distributions, ψ1 and ψ ∞ are 1-norm and ∞-norm respectively; N s is the total number of typical scenarios; T is the scheduling period; ψ pv and ψ wt are the grid connection points of PV and WT respectively; wp is the cost of curtailing wind and solar power; ess is the grid connection point collection of energy storage devices; ess is the unit charging and discharging cost of ESS; δ is the unit cycle cost of ESS; loss Penalty cost for network loss; kg is the cost of opening and closing a single switch; il and ψ tl are the IL and TL load access point sets respectively; il and χ tl are the unit response compensation cost and unit dispatch cost of IL and TL respectively; is the predicted value of the load at node i in period t; ψ B1 , B2 and ψ B3are residential area, commercial area and industrial load nodes respectively; v1 , χ v2 and χ v3 are the unit voltage deviation costs of load nodes in residential, commercial and industrial areas respectively.

[0059] ψ1 and ψ ∞ The constraints are (uncertain set):

[0060]

[0061] {p s The confidence level of} can be expressed as:

[0062]

[0063] A simple conversion gives:

[0064]

[0065] In the formula, θ1 and θ ∞ are the fluctuation values ​​allowed by the probability distribution of actual scenarios under the 1-norm and ∞-norm conditions respectively; α1 and α ∞ is the probability distribution {p s} is the confidence level satisfied; N is the original scene.

[0066] Due to the adoption of the above technical solution, the technical progress achieved by the present invention is:

[0067] The present invention proposes to take the minimum voltage fluctuation penalty cost and operating cost of the active distribution network partition as the objective function, integrates various adjustable resources including distributed power sources, energy storage, various reactive compensation devices, and demand response loads, and on this basis, applies the switch reconstruction as a flexible adjustment resource to the economic, safe, and reliable comprehensive operation of the distribution network, rather than the traditional reduction of network losses and improvement of voltage levels, which is conducive to giving full play to the benefits of switch reconstruction in the comprehensive operation of the active distribution network.

[0068] The present invention considers the differences in demand response regulation among different types of users, and can regulate the demand response levels of various types of users under the premise of limited publicity resources, giving full play to the demand response regulation capability under different types of user participation.

[0069] The uncertainty multi-objective optimization scheduling method of active distribution network considering reconstruction and user participation of demand response proposed in the present invention takes economy and safety as optimization goals under the premise of considering the uncertainty of wind and solar power, gives full play to the inherent advantages of reconstruction and demand response under different types of user participation and the potential for synergy between the two, which is of great significance. BRIEF DESCRIPTION OF THE DRAWINGS

[0070] Figure 1 It is a schematic diagram of the active power distribution network optimization operation method of the present invention;

[0071] Figure 2 It is a schematic diagram of the functional relationship between the participation degree of different types of user demand response and the publicity resources of the present invention;

[0072] Figure 3 It is a flow chart of the algorithm for solving the distributed robust optimization model of the present invention;

[0073] Figure 4 It is a topological structure diagram of the improved IEEE33 node system of the present invention;

[0074] Figure 5 It is the 24h prediction curve diagram of wind, solar and load of the present invention. DETAILED DESCRIPTION

[0075] The present invention is further described in detail below in conjunction with embodiments:

[0076] like Figure 1 As shown, an active distribution network uncertainty optimization method considering reconstruction and user participation includes the following steps:

[0077] Step S1: constructing active distribution network flow constraints including switch reconfiguration based on active distribution network flow constraints.

[0078]

[0079] In the formula, r ij is the equivalent resistance of branch ij; x ij is the equivalent reactance of branch ij; P j,t and Q j,t are the active and reactive power injected by node j during period t; α ij,t is the on-off state variable of branch ij in period t, which is a Boolean variable. The variable value 0 indicates disconnection and 1 indicates closing. ij,t and Q ij,t are the active and reactive power flowing through branch ij during period t respectively; I ij,t is the current value of line ij during period t; V i,t is the voltage amplitude of node i at time period t; P ij,max , Q ij,max ,I ij,max are respectively the square of the current amplitude of branch ij in period t, the square of the voltage amplitude of node i in period t, the maximum active power borne by branch ij, the maximum reactive power borne by branch ij, and the maximum square of the current amplitude borne by branch ij; f(j) is the set of branch head nodes with j as the end node; s(j) is the set of branch end nodes with j as the head node; ψE Represents the set of all branches in the distribution network.

[0080] Step S2: Considering the demand response under different types of user participation based on limited publicity resources, a functional relationship between different types of user participation and demand response publicity resources is established, and a modified demand response constraint condition is constructed.

[0081] like Figure 2 As shown in the figure, it is the functional relationship between different types of user participation and demand response publicity resources, and its expression is:

[0082]

[0083]

[0084] In the formula, x d is the dead zone threshold; x f Saturation threshold; s i is user engagement; x i Promote resources for demand response; 1,dr , ψ 2,dr and ψ 3,dr are the load nodes for demand response in residential areas, commercial areas and industrial areas respectively; x n For fixed publicity resources, it is regarded as a constant.

[0085] When the publicity resources are lower than the minimum dead zone threshold, the user participation is 0. When the publicity resources are greater than the minimum dead zone threshold and less than the minimum saturation threshold, the user participation shows a linear growth trend. When the publicity resources are greater than the minimum saturation threshold, the user participation does not increase with the change of publicity resources and always maintains a value of 1.

[0086] Furthermore, the modified demand response constraints include interruptible load constraints and shiftable load constraints, as follows:

[0087] Interruptible load constraints:

[0088]

[0089] In the formula, and are respectively the active and reactive interruption decision variables of the interruptible load at node i in period t; and are respectively the upper limits of active and reactive interruption of the interruptible load at node i in period t;

[0090] Translational load constraints:

[0091]

[0092]

[0093] In the formula, is the load level of the translatable load at node i during period t; is the total power demand of the shiftable load at node i during the scheduling period; and are the minimum load level and maximum load level of the load that can be translated by node i in period t.

[0094] Step S3: Aiming at the uncertainty of wind and solar power, taking the penalty cost of zone voltage fluctuation and operating cost as the objective function, a multi-scenario distributed robust optimization scheduling model for active distribution network is constructed according to the power flow constraints including switch reconstruction, demand response constraints under different types of user participation, radial structure constraints, safe operation constraints and regulation resource operation constraints.

[0095] Among them, the objective function including the penalty cost of voltage fluctuation in different zones and the operation cost includes two-stage costs. The first-stage cost includes the cost of switch opening and closing and the cost of energy storage cycle. The second-stage cost includes the cost of wind and solar power abandonment, energy storage charging and discharging cost, network loss cost, demand response cost and the cost of voltage fluctuation in different zones. The expression of the objective function is:

[0096]

[0097] In the formula, x is the first-stage variable; y s is the second-stage variable in s scenarios; p s is the probability of each scenario; ψ=ψ1∩ψ ∞ is the set of scene probability distributions, ψ1 and ψ ∞ are 1-norm and ∞-norm respectively; N s is the total number of typical scenarios; T is the scheduling period; ψ pv and ψ wt are the grid connection points of PV and WT respectively; wp is the cost of curtailing wind and solar power; ess is the grid connection point collection of energy storage devices; ess is the unit charging and discharging cost of ESS; δ is the unit cycle cost of ESS; loss Penalty cost for network loss; kg is the cost of opening and closing a single switch; il and ψ tl are the IL and TL load access point sets respectively; il and χ tl are the unit response compensation cost and unit dispatch cost of IL and TL respectively; is the predicted value of the load at node i in period t; ψ B1 , B2 and ψ B3are residential area, commercial area and industrial load nodes respectively; v1 , χ v2 and χ v3 are the unit voltage deviation costs of load nodes in residential, commercial and industrial areas respectively.

[0098] ψ1 and ψ ∞ The constraints are (uncertain set):

[0099]

[0100] {p s The confidence level of} can be expressed as:

[0101]

[0102] A simple conversion gives:

[0103]

[0104] In the formula, θ1 and θ ∞ are the fluctuation values ​​allowed by the probability distribution of actual scenarios under the 1-norm and ∞-norm conditions respectively; α1 and α ∞ is the probability distribution {p s} is the confidence level satisfied; N is the original scene.

[0105] Specifically, in addition to the power flow constraints of switch reconfiguration and the demand response constraints under different types of user participation, other constraints include radial structure constraints, safe operation constraints, and regulation resource operation constraints, among which:

[0106] Radial structure constraints:

[0107]

[0108] In the formula, β ij,t It represents the flow direction of the current in the ij branch during time period t. When the value is 1, it means that the current flows from j to i. H(i) is the set of all nodes adjacent to node i.

[0109] The safe operation constraints include node voltage constraints and branch current constraints, where:

[0110] Node voltage constraints:

[0111]

[0112] Branch current constraints:

[0113]

[0114] Where V i,min and V i,maxare the minimum and maximum voltage allowed at node i respectively.

[0115] The operation constraints of regulating resources include wind power and photovoltaic output constraints, switch opening and closing times constraints, energy storage system operation constraints, discrete reactive power compensation device constraints, and continuous reactive power compensation constraints, among which:

[0116] Wind power and photovoltaic output constraints:

[0117]

[0118] In the formula, and are the wind and solar outputs of node i in period t respectively; and They are the corresponding wind power and photovoltaic predicted power respectively;

[0119] Switch opening and closing times constraints:

[0120]

[0121] In the formula, ψ E is the set of all branches; c max Indicates the total number of times the switch is opened and closed within the cycle; c d,max Indicates the total number of times a single switch is opened and closed within a cycle;

[0122] Energy storage system operation constraints:

[0123]

[0124]

[0125]

[0126] In the formula, and are the charging sign and discharging sign at node i at time t respectively; and are the charging power and discharging power at node i at time t respectively; and They are the upper limits of the active charging and discharging power of ESS respectively; is the ESS capacity at node i during period t; and Respectively represent the minimum and maximum values ​​of the ESS capacitance at node i; η loss , η ch and η dis They represent the self-consumption rate of electric energy, charging efficiency and discharging efficiency respectively.

[0127] Discrete reactive power compensation device constraints:

[0128]

[0129]

[0130]

[0131] In the formula, ψ CB It is the collection of CB grid connection points; is the reactive compensation power of CB at node i at time t; is the number of CB grid-connected groups, which is a discrete variable; is the reactive compensation power of each CB group at node i at time t; and They are grid-connected action identifiers, which are Boolean variables; The maximum number of switching times within a cycle; It is the maximum number of CB groups switched at each time at node i.

[0132] Constraints for continuous reactive power compensation:

[0133]

[0134] In the formula, ψ SVC It is a collection of SVC grid connection points; represents the reactive compensation power of SVC at node i at time t; and They represent the minimum and maximum values ​​of reactive power compensation of SVC at node i respectively.

[0135] Step S4: Use Latin hypercube sampling and improved K-means clustering algorithm to generate multiple typical scenarios, and use column and constraint generation algorithm to solve the model. The solution algorithm flow chart is as follows: Figure 3 As shown, the optimal dispatching scheme of active distribution network is obtained.

[0136] Latin hypercube sampling is a stratified sampling method that allows sampling points to evenly cover all sampling areas. Compared with Monte Carlo sampling, it has a higher sampling rate, higher calculation accuracy and good robustness.

[0137] Improve K-means so that the initial cluster centers are evenly distributed in all scene samples, avoiding the problem of low accuracy caused by randomly selecting cluster centers. The specific steps are as follows:

[0138] Step S401: Assume that the initial number of clusters is k, and randomly select a scene as the initial cluster center;

[0139] Step S402: Calculate the Euclidean distances between the remaining scenes and the cluster center, select the scene with the farthest distance as the second initial cluster center, then calculate the sum of the Euclidean distances between the remaining scenes except the known initial cluster center and the two initial cluster centers, select the scene with the largest sum of distances as the third initial cluster center, and so on, to generate k initial cluster centers;

[0140] Step S403: Calculate the Euclidean distance between each scene and the cluster center, classify each scene into the class where the cluster center with the smallest distance is located, and recalculate the cluster center of each class;

[0141] Step S404: Repeat step S403 until the cluster center does not change or the maximum number of iterations is reached, and the obtained cluster center is the desired typical scene.

[0142] The column and constraint generation algorithm divides the model into a master-problem (MP) and a sub-problem (SP) and solves them iteratively. Compared with the traditional Benders decomposition algorithm, it has the advantages of fewer iterations and higher accuracy.

[0143] Main question:

[0144] Given the probability distribution of each scenario (the initial probability distribution of each clustered scenario is used for the first solution, and the probability distribution of each scenario under the worst case obtained by the sub-problem is used for subsequent solutions), find the optimal solution that satisfies the lowest overall operating cost of the system. The main problem is as follows:

[0145] MP:

[0146]

[0147] In the formula, τ is an intermediate variable, which represents the estimated value of the sub-problem; is the worst scenario probability distribution obtained for the mth iteration subproblem; m is the current iteration number, and M is the total number of model iterations.

[0148] By solving the main problem MP, we can get the optimal solution variable x * and a lower bound for the model.

[0149] Sub-questions:

[0150] The sub-problem is a max-min double-layer structure. The sub-problem is given the first-stage variable x * Based on this, the probability distribution of the worst scenario is obtained and passed to the main problem and an upper limit value of the model. The sub-problem form is as follows:

[0151] SP:

[0152] Since the inner and outer variables of the sub-problem are unrelated, they can be decomposed into two steps for solution. First, solve the inner min problem and then solve the outer max problem. The inner SP1 solves the optimal objective function value and the optimal second-stage variable, and the outer SP2 solves the worst probability distribution corresponding to the maximum objective function in the probability distribution set. The specific form is as follows:

[0153] SP1:

[0154] SP2:

[0155] The present invention is further described below in conjunction with specific embodiments.

[0156] The present invention uses the improved IEEE33 node system as an example to verify the effectiveness of the present invention. The node system includes an energy storage device ESS, a wind and solar distributed power source (WT, PV), a controllable switch, a variety of reactive power compensation devices (CB, SVC), and an interruptible load (IL) and a shiftable load (TL). The specific system structure diagram is as follows Figure 4 As shown in the figure, the charging and discharging power limit of ESS is [0, 0.4MW], the initial power is set to 0.35MWh, the self-loss rate is 0.005, the charging and discharging efficiency is 0.9 and 1.11, and the ESS capacity limit is [0.2MWh, 1.5MWh]; the 24h curve prediction diagram of WT and PV is shown in the figure Figure 5 As shown, the prediction error follows a normal distribution with a mean of 0 and a standard deviation of 0.2 times the prediction value; the upper limit of the number of switch actions within the controllable switch scheduling period is 24, and the upper limit of the number of actions of a single switch within the scheduling period is 5; there are 5 groups of CBs, each with 0.1Mvar, the total number of switching times within the scheduling period is upper limit 5 times, and the upper limit of the number of CB groups each time the CB is switched is 5; the reactive output limit of SVC is [-0.2Mvar, 0.3Mvar]; the upper limit of the active and reactive response of IL at each moment is 0.15 times the node load at that moment; the response range of TL is 0.75 times the node load at that moment to 1.25 times the node load at that moment; refer to Table 1 for each cost type; refer to Time-of-Use Electricity Price Table 2 for unit network loss cost. The system sets the reference voltage to 12.66KV, the reference power to 1MVA, the constraint range of the line node voltage to 0.94-1.06, the commercial loads to nodes: 2, 3, 12, 13, 18, 19, 20, 21, 22, 28, 30, 32, the industrial loads to nodes: 6, 7, 23, 24, 29, 31, and the residential loads to nodes: 4, 5, 8, 9, 10, 11, 14, 15, 16, 17, 25, 26, 27, 32. The model is established using MATLAB, and the commercial solver CPLEX is called for solution. The example sets up 7 scenarios to analyze the effectiveness of the method of the present invention.

[0157] Scenario 1: Without considering reconstruction and considering the equal distribution of publicity resources in response to different types of user needs;

[0158] Scenario 2: Considering reconstruction and even distribution of publicity resources in response to different types of user needs;

[0159] Scenario 3: No consideration of reconstruction and optimal allocation of publicity resources in response to different types of user needs;

[0160] Scenario 4: Considering reconstruction and optimizing the allocation of publicity resources in response to different types of user needs;

[0161] Scenario 5: Based on Scenario 4, further explore the unification of voltage fluctuation costs for different users;

[0162] Scenario 6: Based on Scenario 4, further explore random optimization;

[0163] Scenario 7: Based on Scenario 4, further explore robust optimization.

[0164] Table 1

[0165]

[0166]

[0167] Table 2

[0168] type Time Electricity price (yuan / Mwh) Peak electricity price 9~11,19~23 1350 Normal electricity price 8,12~18 900 Off-peak electricity price 1~7,24 480

[0169] Table 3

[0170]

[0171]

[0172] Table 3 is a comparison of scenarios 1 to 4. By comparing scenarios 1 and 2, we can see that the network loss is significantly reduced, the new energy consumption is increased by 9%, the overall voltage fluctuation is lower, and the comprehensive cost is reduced by 358.7. Therefore, considering the switch reconstruction can significantly improve the system economy, increase the new energy consumption, and reduce voltage fluctuations. By comparing scenarios 1 and 3, we can see that considering the optimal allocation of publicity resources can also improve the system economy, the new energy consumption rate, and reduce voltage fluctuations. By comparing scenarios 1 and 4, we can see that under comprehensive coordination, the network loss can be reduced more effectively, the new energy consumption rate is increased by 14%, the voltage fluctuation is significantly reduced, and the comprehensive cost is reduced by 8%.

[0173] Table 4

[0174]

[0175] Table 4 is a comparison of the comprehensive costs of scenario 4 and scenario 5. It can be seen from the table that, considering the differences in voltage quality requirements in different user areas, the partitioned voltage fluctuation cost can effectively improve the system's scheduling potential and reduce the comprehensive cost of system operation.

[0176] Table 5

[0177]

[0178] Table 5 is a comprehensive cost comparison of scenarios 4, 6, and 7. It can be seen from the table that the robust optimization not only considers the probability distribution of uncertain variables but also the probability distribution of the worst scenario based on the robust idea, taking into account both economy and robustness.

[0179] The results of the example show that the method proposed in this patent takes into account the uncertainty of wind and solar power, and takes the economic operation cost and the partition voltage fluctuation cost as the objective function, giving full play to the inherent advantages of reconstruction and demand response under different types of user participation and the potential for synergy between the two, effectively improving the economy and safety of active distribution network operation.

[0180] It is to be understood that the present invention is described by some embodiments, and it is known to those skilled in the art that various changes or equivalent substitutions may be made to these features and embodiments without departing from the spirit and scope of the present invention. In addition, under the teachings of the present invention, these features and embodiments may be modified to adapt to specific circumstances and materials without departing from the spirit and scope of the present invention. Therefore, the present invention is not limited by the specific embodiments disclosed herein, and all embodiments falling within the scope of the claims of this application are within the scope of protection of the present invention.

Claims

1. An active distribution network uncertainty optimization method considering reconstruction and user participation, characterized by: The following steps are involved: Step S1: construct active distribution network flow constraints including switch reconstruction based on active distribution network flow constraints, the constraints are: In the formula, r ij is the equivalent resistance of branch ij; x ij is the equivalent reactance of branch ij; P j,t and Q j,t are the active and reactive power injected by node j during period t; α ij,t is the on-off state variable of branch ij in period t, which is a Boolean variable. The variable value 0 indicates disconnection and 1 indicates closing. ij,t and Q ij,t are the active and reactive power flowing through branch ij during period t respectively; I ij,t is the current value of line ij during period t; V i,t is the voltage amplitude of node i at time period t; P ij,max , Q ij,max ,I ij,max are respectively the square of the current amplitude of branch ij in period t, the square of the voltage amplitude of node i in period t, the maximum active power borne by branch ij, the maximum reactive power borne by branch ij, and the maximum square of the current amplitude borne by branch ij; f(j) is the set of branch head nodes with j as the end node; s(j) is the set of branch end nodes with j as the head node; ψ E Represents the set of all branches in the distribution network; Step S2: Considering the demand response under different types of user participation based on limited publicity resources, a functional relationship between different types of user participation and demand response publicity resources is established, and a modified demand response constraint condition is constructed. The functional relationship is: In the formula, x d is the dead zone threshold; x f Saturation threshold; s i is user engagement; x i Promote resources for demand response; 1,dr , ψ 2,dr and ψ 3,dr are the load nodes for demand response in residential areas, commercial areas and industrial areas respectively; x n For fixed publicity resources, it is considered as a constant; When the publicity resources are lower than the minimum dead zone threshold, the user participation is 0. When the publicity resources are greater than the minimum dead zone threshold and less than the minimum saturation threshold, the user participation shows a linear growth trend. When the publicity resources are greater than the minimum saturation threshold, the user participation does not increase with the change of publicity resources and always maintains a value of 1. The modified demand response constraint condition includes an interruptible load constraint and a shiftable load constraint, wherein: Interruptible load constraints: In the formula, and are respectively the active and reactive interruption decision variables of the interruptible load at node i in period t; and are respectively the upper limits of active and reactive interruption of the interruptible load at node i in period t; Translational load constraints: In the formula, is the load level of the translatable load at node i during period t; is the total power demand of the shiftable load at node i during the scheduling period; and is the minimum load level and maximum load level of the load that can be translated by node i in period t; Step S3: In view of the uncertainty of wind and solar power, the penalty cost of voltage fluctuation in different zones and the operating cost are taken as the objective function, reactive support and energy storage are added, and a multi-scenario distributed robust optimization dispatch model of active distribution network is constructed according to the power flow constraints including switch reconstruction, demand response constraints under different types of user participation, radial structure constraints, safe operation constraints and regulation resource operation constraints; The objective function including the penalty cost of voltage fluctuation in different zones and the operation cost includes two-stage costs. The first-stage cost includes the cost of switch opening and closing and the cost of energy storage cycle. The second-stage cost includes the cost of wind and solar power abandonment, energy storage charging and discharging cost, network loss cost, demand response cost and the cost of voltage fluctuation in different zones. The objective function expression is: In the formula, x is the first-stage variable; y s is the second-stage variable in the sth scenario; p s is the probability of each scenario; ψ=ψ1∩ψ ∞ is the set of scene probability distributions, ψ1 and ψ ∞ are 1-norm and ∞-norm respectively; N s is the total number of typical scenarios; T is the scheduling period; ψ pv and ψ wt are the grid connection points of PV and WT respectively; wp is the cost of curtailing wind and solar power; ess is the grid connection point collection of energy storage devices; ess is the unit charging and discharging cost of ESS; δ is the unit cycle cost of ESS; loss Penalty cost for network loss; kg is the cost of opening and closing the switch once; ψ il and ψ tl are the IL and TL load access point sets respectively; il and χ tl are the unit response compensation cost and unit dispatch cost of IL and TL respectively; is the predicted value of the load at node i in period t; ψ B1 , B2 and ψ B3 are residential area, commercial area and industrial load nodes respectively; v1 , χ v2 and χ v3 are the unit voltage deviation costs of load nodes in residential, commercial and industrial areas respectively; ψ1 and ψ ∞ The constraints are uncertain sets: {p s The confidence level of} can be expressed as: A simple conversion gives: In the formula, θ1 and θ ∞ are the fluctuation values ​​allowed by the probability distribution of actual scenarios under the 1-norm and ∞-norm conditions respectively; α1 and α ∞ is the probability distribution {p s }The confidence level satisfied; N is the original scene; Step S4: Use Latin hypercube sampling and improved K-means clustering algorithm to generate multiple typical scenarios, and use column and constraint generation algorithm to solve the model to obtain the active distribution network optimization scheduling plan.

2. The active distribution network uncertainty optimization method considering reconstruction and user participation according to claim 1 is characterized by: In step S3, the radial structure constraint is: In the formula, β ij,t It represents the flow direction of the current in the ij branch during time period t. When the value is 1, it means that the current flows from j to i. H(i) is the set of all nodes adjacent to node i.

3. The active distribution network uncertainty optimization method considering reconstruction and user participation according to claim 2 is characterized in that: In step S3, the safety operation constraints include node voltage constraints and branch current constraints, wherein: Node voltage constraints: Branch current constraints: Where V i,min and V i,max are the minimum and maximum voltage allowed at node i respectively.

4. The active distribution network uncertainty optimization method considering reconstruction and user participation according to claim 3 is characterized by: In step S3, the regulating resource operation constraints include wind power and photovoltaic output constraints, switch opening and closing times constraints, energy storage system operation constraints, discrete reactive compensation device constraints, and continuous reactive compensation constraints, wherein: Wind power and photovoltaic output constraints: In the formula, and are the wind and solar outputs of node i in period t respectively; and They are the corresponding wind power and photovoltaic predicted power respectively; Switch opening and closing times constraints: In the formula, ψ E is the set of all branches; c max Indicates the total number of times the switch is opened and closed within the cycle; c d,max Indicates the total number of times a single switch is opened and closed within a cycle; Energy storage system operation constraints: In the formula, and are the charging sign and discharging sign at node i at time t respectively; and are the charging power and discharging power at node i at time t respectively; and They are the upper limits of the active charging and discharging power of ESS respectively; is the ESS capacity at node i during period t; and Respectively represent the minimum and maximum values ​​of the ESS capacitance at node i; η loss , η ch and η dis They represent the electric energy self-consumption rate, charging efficiency and discharging efficiency respectively; Discrete reactive power compensation device constraints: In the formula, ψ CB It is the collection of CB grid connection points; is the reactive compensation power of CB at node i at time t; is the number of CB grid-connected groups, which is a discrete variable; is the reactive compensation power of each CB group at node i at time t; and They are grid-connected action identifiers, which are Boolean variables; The maximum number of switching times within a cycle; is the maximum number of CB switching groups at each time at node i; Constraints for continuous reactive power compensation: In the formula, ψ SVC It is a collection of SVC grid connection points; represents the reactive compensation power of SVC at node i at time t; and They represent the minimum and maximum values ​​of reactive power compensation of SVC at node i respectively.