Dynamic robust reconstruction method of distribution network based on mobile energy storage under non-ideal conditions
By establishing a mobile energy storage model and an ellipsoid uncertainty set in the distribution network in a non-ideal state, combining the dynamic reconstruction model of the AC-DC hybrid distribution network, and using the C&CG algorithm, the distribution network is solved in the non-ideal state. Dynamic robust reconstruction problem achieves reduced energy loss and improved robustness.
Patent Information
- Application Number
- CN202411011103.8
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Filing Date
- 2024-07-26
- Publication Date
- 2024-11-12
- Estimated Expiration
- 2044-07-26
AI Technical Summary
In the face of fluctuations in renewable energy access and power demand, it is difficult to achieve dynamic robust reconstruction. Especially in the non-ideal state, the model of mobile energy storage assumes that the charging and discharging power is constant and cannot effectively reduce energy. Loss and improve robustness.
By establishing a mobile energy storage model based on non-ideal state, the operating area of charge and discharge power is sampled and converted into a linear model, and an ellipsoid uncertain set is established using the historical data of new energy output to build a data-driven convex hull polyhedron to build a data-driven convex hull polyhedron Set, combined with the dynamic reconstruction model of AC and DC hybrid distribution network, and the solution is performed using C&CG algorithm.
Realize dynamic robust reconstruction of the distribution network under non-ideal state, reduces energy loss, improves the economy and practicality of the model, reduces the conservatism of the results, enhances robustness, and adapts to distributed photovoltaics. Volatility and randomness of output.
Smart Images

Figure CN118539439B_ABST
Abstract
Description
Technical Field
[0001] The present invention belongs to the technical field of smart grids, and in particular relates to a method for dynamic robust reconstruction of a distribution network based on mobile energy storage under non-ideal conditions. Background Art
[0002] As a bridge between power generation and end users, the role of distribution network is becoming increasingly significant. However, traditional distribution network faces considerable challenges in the face of widespread access to renewable energy and growing power demand. At present, technological innovation and structural adjustment of distribution network have become the key to ensure the sustainable development of power system.
[0003] From the perspective of technological innovation, firstly, by adopting AC / DC hybrid technology, electric energy can be transmitted more efficiently, and energy loss can be effectively reduced. In addition, the stability and flexibility of the power grid can be improved, which makes it possible to access more renewable energy generation. Secondly, in terms of load, the development of demand-side response enables consumers to actively adjust their electricity use according to power market signals or power grid demand, thereby participating in the balance and optimization of the power grid. This participation not only reduces the burden on the power grid, but also provides consumers with opportunities for energy saving and cost reduction. Finally, in terms of energy storage, the progress of energy storage technology not only greatly enhances the flexibility and stability of the power system, but also provides frequency regulation, emergency standby, power quality management and other services, becoming an important tool for power grid regulation. Some studies have established energy storage-type intelligent soft switch models to improve the operation level of distribution networks, and proposed measurement-based energy storage models to avoid the shortcomings of traditional energy storage models. Some scholars have adopted mobile energy storage strategies to give full play to their spatial and temporal support role. Finally, the effectiveness of the model has been verified by a 33-node distribution network.
[0004] From the perspective of structural adjustment, in order to more directly and effectively reduce the energy loss of the distribution network within a limited time, network reconstruction is often required. According to different time scales, distribution network reconstruction can be divided into two forms: static and dynamic. Dynamic reconstruction has attracted much attention because it can perform real-time dynamic optimization of the distribution network. Some literature first divides the dynamic reconstruction period into sections and then performs static reconstruction. Then, the static reconstruction results of each time period are compared, such as network loss and switching status. Similar time periods are then merged to reduce unnecessary switching operations. Some scholars use a two-layer dynamic time period division method. At the upper level, the switch state quantity of each time period is determined with the goal of minimizing the sum of network loss cost and wind and solar power abandonment costs. At the lower level, the goal is to minimize network loss.
[0005] As the penetration rate of distributed generation (DG) such as photovoltaic and wind power in distribution networks increases, the reconstruction of distribution networks considering the uncertainty factors brought by DG has important practical significance. There are two main research methods to solve this uncertainty at this stage, one is the stochastic optimization method and the other is the robust optimization method. Compared with the stochastic optimization method, the robust optimization method reduces the dimension of the solution, so it has received widespread attention.
[0006] In order to enhance the reliability of robust optimization results and describe the correlation between uncertain parameters, some scholars have used historical data of uncertain variables to try to find the connection between the changes of random variables in recent years, and proposed data-driven uncertainty sets. This model introduces the concept and construction method of dynamic uncertainty sets to model the spatiotemporal correlation of uncertainty; based on the polyhedron set, the boundary of the polyhedron set is bent using mathematical analysis methods to establish a correlation polyhedron set model; some scholars have further improved and constructed a generalized correlation polyhedron set model so that the polyhedron set can better envelop the range of uncertain parameters.
[0007] In addition to using polyhedron sets and ellipse sets for construction, another commonly used method is to construct uncertain sets based on extreme scenarios. Some scholars first select historical data of uncertain sets, then filter out extreme scenarios based on historical data, and construct convex hull sets based on these scenarios, and introduce appropriate scaling factors to cover all historical data, and finally establish a robust optimization model based on extreme scenarios. Summary of the invention
[0008] The technical problem to be solved by the present invention is to provide a dynamic robust reconstruction method of a distribution network based on mobile energy storage under non-ideal conditions, which is used for modeling and robust optimization of AC / DC hybrid distribution networks.
[0009] The technical solution adopted by the present invention to solve the above technical problems is: a method for dynamic robust reconstruction of a distribution network based on mobile energy storage under non-ideal conditions, comprising the following steps:
[0010] S1: Establish the relationship between the upper and lower limits of charge and discharge power and the charge and discharge current, the input power of the battery and the output power of the battery from the charge and discharge process inside the battery; sample the operation area of the charge and discharge power, and transform the reconstructed battery energy storage model into a linear model in the sampling point set of the convex envelope;
[0011] S2: Select the minimization of the comprehensive cost within the dispatching period as the objective function and establish a dynamic reconstruction model of the AC / DC hybrid distribution network;
[0012] S3: Use the historical data of renewable energy output to establish an ellipsoid uncertainty set, and establish a data-driven convex hull polyhedron set by connecting the vertices of the high-dimensional ellipsoid;
[0013] S4: The model established in step S2 is converted into a matrix form and the C&CG algorithm is used to solve the model.
[0014] According to the above scheme, in step S1, for the node Department The battery energy storage connected during the time period is set is the input power of the battery, is the output power of the battery; and is a variable associated with the charge and discharge sample data; for the discharge curve, let , , They represent the battery energy storage discharge sample sampling, J is the sampling point set of the discharge sample data; for the charging curve, let , , They represent the battery energy storage charging sample sampling, and K is the sampling point set of the charging sample data; the battery energy storage model under non-ideal conditions is expressed as a linear combination of sample sampling data:
[0015] ,
[0016] ,
[0017] ,
[0018] ,
[0019] ,
[0020] ,
[0021] ,
[0022] .
[0023] Furthermore, in step S2, the network loss cost coefficient is used The switching operation cost coefficient is expressed as The cost coefficient of abandoned light is expressed as The power purchase cost coefficient is expressed as express; is a Boolean variable representing Branch circuit in the distribution network system The breaking condition of the branch ,definition Time from the beginning Flow to the end The power flow is , from the end Flow to the head end The power flow is ; , is with The physical quantities related to the output of the distributed photovoltaic model at each moment are the available power and the actual output power after the distributed photovoltaic fluctuations; for The power purchased at the time limit; represents the comprehensive cost function of distribution network reconstruction; Represents the set of all branches of the distribution network; represents the time set of the distribution network reconstruction; Represents the node set of distributed photovoltaic access; for the dynamic reconstruction model of AC / DC hybrid distribution network, the minimization of the comprehensive cost within the dispatching period is selected as the objective function:
[0024] ,
[0025] The first term in the above formula represents the system network loss cost, the second term represents the system switching action cost, the third term represents the system abandoned light cost, and the fourth term represents the battery energy storage aging cost.
[0026] Furthermore, in step S2, the constraints include power flow constraints, on-load tap-changing transformer constraints, radial topology, AC / DC converter station constraints, capacitor constraints, distributed photovoltaic operation constraints, mobile energy storage operation constraints, demand-side response model and switch constraints. The specific calculation steps are as follows:
[0027] S21: Pair Node exist The power flow variables at time are The total active power injected into the node, The total reactive power injected into the node, Power injected into mobile energy storage, Active load after applying demand-side response mechanism to the node, is the reactive load of the node, The reactive power injected into the gateway, is the reactive power generated by the capacitor; definition Time from the beginning Flow to the end The reactive power is , from the end Flow to the head end The reactive power is ; For all nodes A set of connected nodes; for branches ,set up is the branch current amplitude, , is the physical parameter of the branch; definition , , ,in, , is the voltage across the branch ij, is the voltage phase difference at both ends, and the variable The variables related to the branch switch state and the branch terminal voltage are expressed as , and , The relationship between them; then the power flow constraint is:
[0028] ,
[0029] ,
[0030] ,
[0031] ,
[0032] ,
[0033] ,
[0034] ,
[0035] ,
[0036] ,
[0037] ;
[0038] Set Node The maximum voltage at , the minimum value is , add additional constraints as follows:
[0039] ,
[0040] ;
[0041] Set up branch road The maximum current is , the branch current constraint is:
[0042] ;
[0043] S22: Set is the ratio parameter, is the gear number parameter, is the gear interval value parameter, is the upper limit parameter of the ratio, is the lower limit parameter of the transformation ratio; is an integer variable representing the transformer gear position; , , ; Then the voltage relationship between the two ends of the transformer, that is, the on-load tap-changing transformer constraint is:
[0044] ,
[0045] ,
[0046] ,
[0047] ;
[0048] S23: Set Represents a set of distribution network nodes. Represents the root node set; , Both are Boolean variables used to judge nodes and nodes The relationship between them; then the radial topology is:
[0049] ;
[0050] S24: In the AC / DC converter station model, , They are The power flow measured by the AC converter station at every moment; for The DC power flow is measured at the converter station at all times; is the reference point voltage, let , ; is a set of converter branches; Indicates the DC side voltage utilization parameter; represents the modulation parameter; represents the conversion rate parameter; then the power flow equation of the AC / DC converter station model, i.e., the AC / DC converter station constraint, is:
[0051] ,
[0052] ,
[0053] ,
[0054] ,
[0055] ,
[0056] ,
[0057] ,
[0058] ;
[0059] S25: For nodes exist The capacitor bank connected at the moment is set A node assembly for installing capacitors; is the final number of grid-connected capacitor banks; The upper limit of the number of capacitor groups is is the capacity of a single group of capacitors, is the maximum value of the change in the number of capacitor groups during the interval period; then the capacitor constraint is:
[0060] ,
[0061] ;
[0062] S26: For nodes exist Distributed photovoltaic connected at all times, is the maximum available power of distributed photovoltaic before fluctuation; is the maximum fluctuation of distributed photovoltaic, For distributed photovoltaic Uncertain variables at the moment; For distributed photovoltaic Reactive power at the moment, is the capacity of distributed photovoltaics; then the distributed photovoltaic operation constraints are:
[0063] ,
[0064] ,
[0065] ;
[0066] S27: Set It is a Boolean variable, indicating that the mobile storage Whether the moment is docked at the node ; represents the set of all dockable nodes of mobile energy storage; based on the battery energy storage model established in step S1, the mobile energy storage operation constraints are established as follows:
[0067] ,
[0068] ,
[0069] ;
[0070] S28: The demand-side response mechanism applied to the node i load at time t, Represents the set of nodes that perform demand-side response; and are the load change and electricity price change before and after the demand side response is applied, is the electricity price variable after implementing demand response; is the electricity price elasticity coefficient, is the electricity price parameter before applying demand-side response, and are the upper and lower limits of the electricity price after applying demand-side response, Representation Node exist The active load at the moment; then the demand side response model is:
[0071] ,
[0072] ,
[0073] ,
[0074] ;
[0075] S29: Set It is the initial branch state of the distribution network system; It is the upper limit of branch switch action; is the total number of branch switches; then the switch constraint is:
[0076] .
[0077] Furthermore, in step S3, according to the uncertainty variable contained in the constraint of step S2 ,set up represents the center point of the high-dimensional ellipsoid, is a positive definite matrix, which indicates the offset direction of the high-dimensional ellipsoid relative to the coordinate axis; the budget uncertainty set is used for the ellipsoid set The specific expression for describing the fluctuation range of distributed photovoltaic output is:
[0078] .
[0079] Furthermore, in step S3, a data-driven convex hull uncertainty set is proposed, and the specific steps are as follows:
[0080] S31: Construct a high-dimensional ellipsoid uncertainty set with the smallest volume and covering all historical data fluctuations ;
[0081] S32: Based on the original high-dimensional ellipsoid, the positive definite matrix Perform orthogonal decomposition and rotate and translate the original ellipsoid so that the center of the ellipsoid falls on the center of the coordinate axis. At this time, the high-dimensional ellipsoid uncertainty set is ;
[0082] S33: Scaling the high-dimensional linear polyhedron set in step S2 so that there is a minimum Make the scaled polyhedron set just envelop all possible data points, The value range is ;
[0083] set up Indicates The weight coefficient of each vertex; is the scaling factor, which is used to adjust the conservativeness of the envelope range of the high-dimensional linear polyhedron; then the uncertain set of the high-dimensional linear polyhedron after scaling is for:
[0084] ;
[0085] S34: Rotate and translate the scaled high-dimensional linear polyhedron to fit the original data point range; then the high-dimensional linear polyhedron after rotation and translation is an uncertain set for:
[0086] .
[0087] Furthermore, in step S4, the constraints in step S2 are written in matrix form, which are:
[0088] Power flow constraints: ;
[0089] Distributed photovoltaic operation constraints: ;
[0090] On-load tap-changing transformer constraints: ;
[0091] AC / DC converter station constraints: ;
[0092] Demand-side response constraints: ;
[0093] Battery storage constraints: ;
[0094] Constraints for switching capacitors in groups:
[0095] set up represents the decision variable in the first stage, represents the decision variable in the second stage, Indicates uncertain variables; , In the objective function , The relevant parameter matrix; , , , , , , , represents the parameter matrix related to the two-stage robust optimization constraints; Indicates a given When continuous variable The feasible domain of the first stage objective function represents the switching cost, the objective function of the second stage Represents the sum of the remaining costs except the switch cost; combined with the objective function, the matrix form of the dynamic robust reconstruction model of the two-stage AC / DC hybrid distribution network is established as follows:
[0096] .
[0097] Furthermore, in the step S4, the C&CG method is used to transform the dynamic robust reconstruction model of the two-stage AC / DC hybrid distribution network into a master-slave problem for solution;
[0098] set up , , is the dual variable constrained by the subproblem, and the corresponding master-slave problem model is:
[0099] ,
[0100] ;
[0101] The outer approximation is used to solve the existing bilinear terms ( ) problem, according to the duality theorem, the inner layer minimization problem of SP1 is changed into the same maximization problem as the outer layer:
[0102] ;
[0103] set up is the number of historical iterations, is the current iteration number, As auxiliary variables, MP2 for solving the upper bound and SP2 for solving the lower bound are:
[0104] ,
[0105] ;
[0106] In the above formula The linearization formula is:
[0107] .
[0108] A distribution network dynamic robust reconstruction system based on mobile energy storage under non-ideal conditions, the system comprising a processor and a memory, the memory storing computer instructions, the processor being used to execute the computer instructions stored in the memory, and when the computer instructions are executed by the processor, the system implements the steps of a distribution network dynamic robust reconstruction method based on mobile energy storage under non-ideal conditions.
[0109] A computer memory stores a computer program executable by a computer processor, wherein the computer program executes a method for dynamic robust reconstruction of a distribution network based on mobile energy storage under non-ideal conditions.
[0110] The beneficial effects of the present invention are:
[0111] 1. The dynamic robust reconstruction method of the distribution network based on mobile energy storage under non-ideal conditions of the present invention is aimed at the traditional energy storage model assuming that the charging and discharging power and efficiency are constant, which is inconsistent with the nonlinear characteristics of the battery unit, and at the same time, the uncertainty modeling of the random disturbance of distributed energy has the disadvantage of being conservative. First, the operating area of the actual battery charging and discharging power is sampled, and the energy storage model based on the measurement is reconstructed into a linear model in the sampling point set, and then a mobile energy storage model based on non-ideal conditions is built to make up for the shortcomings of poor economy and low practicality brought by traditional fixed energy storage; secondly, the historical data of new energy output is used to establish an ellipsoid uncertainty set, and a data-driven convex hull polyhedron set is established by connecting the vertices of the high-dimensional ellipsoid; finally, the established dynamic robust reconstruction model of the AC / DC hybrid distribution network is written in matrix form, and the C&CG algorithm is used to solve it; the functions of AC / DC hybrid distribution network modeling and robust optimization are realized.
[0112] 2. The battery energy storage model under non-ideal conditions of the present invention avoids the shortcomings of traditional models. More importantly, it does not need to introduce binary variables to characterize the charging and discharging status, thereby greatly simplifying the complexity of the model. This simplification makes the battery energy storage model easier to embed into the AC / DC distribution network reconstruction model; at the same time, compared with fixed energy storage, mobile energy storage can make timely adjustments to changes in power flow due to its spatial mobility. Therefore, compared with traditional fixed energy storage, the mobile energy storage model under non-ideal conditions has good economy and practicality.
[0113] 3. The present invention takes into account the volatility and randomness of distributed photovoltaic output. Since the modeling of the three traditional uncertain sets has the disadvantages of high conservatism and poor robustness, compared with the box set, under the condition of the same robust adjustment coefficient, the present invention adopts an adaptive data-driven convex hull set to more accurately describe the distribution range of uncertain parameters; compared with the box set, the present invention reduces the conservatism of the results to a certain extent, while enhancing the robustness. BRIEF DESCRIPTION OF THE DRAWINGS
[0114] Figure 1 The battery storage energy of the embodiment of the present invention is and Discharge curve diagram of the operating area.
[0115] Figure 2 The battery storage energy of the embodiment of the present invention is and Charging curve diagram of the operating area.
[0116] Figure 3 is a flow chart of an embodiment of the present invention. DETAILED DESCRIPTION
[0117] The present invention will be further described in detail below in conjunction with the accompanying drawings and specific embodiments. The specific examples described herein are only used to explain the present invention and are not used to limit the present invention.
[0118] Example 1
[0119] See also Figure 3 , this embodiment includes the following steps:
[0120] The method for dynamic robust reconstruction of distribution network based on mobile energy storage under non-ideal conditions includes the following steps:
[0121] S1: Establish the relationship between the upper and lower limits of charge and discharge power and the charge and discharge current, the input power of the battery and the output power of the battery from the charge and discharge process inside the battery; sample the operation area of the charge and discharge power, and transform the reconstructed battery energy storage model into a linear model in the sampling point set of the convex envelope;
[0122] For nodes Department The battery energy storage connected during the time period is set is the input power of the battery, is the output power of the battery; and is a variable associated with the charge and discharge sample data; for the discharge curve, let , , They represent the battery energy storage discharge sample sampling, J is the sampling point set of the discharge sample data; for the charging curve, let , , They represent the battery energy storage charging sample sampling, and K is the sampling point set of the charging sample data; the battery energy storage model under non-ideal conditions is expressed as a linear combination of sample sampling data:
[0123] ,
[0124] ,
[0125] ,
[0126] ,
[0127] ,
[0128] ,
[0129] ,
[0130] .
[0131] S2: Select the minimization of the comprehensive cost within the dispatching period as the objective function and establish a dynamic reconstruction model of the AC / DC hybrid distribution network;
[0132] The network loss cost coefficient is The switching operation cost coefficient is expressed as The cost coefficient of abandoned light is expressed as The power purchase cost coefficient is expressed as express; is a Boolean variable representing Branch circuit in the distribution network system The breaking condition of the branch ,definition Time from the beginning Flow to the end The power flow is , from the end Flow to the head end The power flow is ; , is with The physical quantities related to the output of the distributed photovoltaic model at each moment are the available power and the actual output power after the distributed photovoltaic fluctuations; for The power purchased at the time limit; represents the comprehensive cost function of distribution network reconstruction; Represents the set of all branches of the distribution network; represents the time set of the distribution network reconstruction; Represents the node set of distributed photovoltaic access; for the dynamic reconstruction model of AC / DC hybrid distribution network, the minimization of the comprehensive cost within the dispatching period is selected as the objective function:
[0133] ,
[0134] The first term in the above formula represents the system network loss cost, the second term represents the system switching action cost, the third term represents the system abandoned light cost, and the fourth term represents the battery energy storage aging cost.
[0135] The constraints include power flow constraints, on-load tap-changing transformer constraints, radial topology, AC / DC converter station constraints, capacitor constraints, distributed photovoltaic operation constraints, mobile energy storage operation constraints, demand-side response model and switch constraints. The specific calculation steps are as follows:
[0136] S21: Pair Node exist The power flow variables at time are The total active power injected into the node, The total reactive power injected into the node, Power injected into mobile energy storage, Active load after applying demand-side response mechanism to the node, is the reactive load of the node, The reactive power injected into the gateway, is the reactive power generated by the capacitor; definition Time from the beginning Flow to the end The reactive power is , from the end Flow to the head end The reactive power is ; For all nodes A set of connected nodes; for branches ,set up is the branch current amplitude, , is the physical parameter of the branch; definition , , ,in, , is the voltage across the branch ij, is the voltage phase difference at both ends, and the variable The variables related to the branch switch state and the branch terminal voltage are expressed as , and , The relationship between them; then the power flow constraint is:
[0137] ,
[0138] ,
[0139] ,
[0140] ,
[0141] ,
[0142] ,
[0143] ,
[0144] ,
[0145] ,
[0146] ;
[0147] Set Node The maximum voltage at , the minimum value is , add additional constraints as follows:
[0148] ,
[0149] ;
[0150] Set up branch road The maximum current is , the branch current constraint is:
[0151] ;
[0152] S22: Set is the ratio parameter, is the gear number parameter, is the gear interval value parameter, is the upper limit parameter of the ratio, is the lower limit parameter of the transformation ratio; is an integer variable representing the transformer gear position; , , ; Then the voltage relationship between the two ends of the transformer, that is, the on-load tap-changing transformer constraint is:
[0153] ,
[0154] ,
[0155] ,
[0156] ;
[0157] S23: Set Represents a set of distribution network nodes. Represents the root node set; , Both are Boolean variables used to judge nodes and nodes The relationship between them; then the radial topology is:
[0158] ;
[0159] S24: In the AC / DC converter station model, , They are The power flow measured by the AC converter station at every moment; for The DC power flow is measured at the converter station at all times; is the reference point voltage, let , ; is a set of converter branches; Indicates the DC side voltage utilization parameter; represents the modulation parameter; represents the conversion rate parameter; then the power flow equation of the AC / DC converter station model, i.e., the AC / DC converter station constraint, is:
[0160] ,
[0161] ,
[0162] ,
[0163] ,
[0164] ,
[0165] ,
[0166] ,
[0167] ;
[0168] S25: For nodes exist The capacitor bank connected at the moment is set A node assembly for installing capacitors; is the final number of grid-connected capacitor banks; The upper limit of the number of capacitor groups is is the capacity of a single group of capacitors, is the maximum value of the change in the number of capacitor groups during the interval period; then the capacitor constraint is:
[0169] ,
[0170] ;
[0171] S26: For nodes exist Distributed photovoltaic connected at all times, is the maximum available power of distributed photovoltaic before fluctuation; is the maximum fluctuation of distributed photovoltaic, For distributed photovoltaic Uncertain variables at the moment; For distributed photovoltaic Reactive power at the moment, is the capacity of distributed photovoltaics; then the distributed photovoltaic operation constraints are:
[0172] ,
[0173] ,
[0174] ;
[0175] S27: Set It is a Boolean variable, indicating that the mobile storage Whether the moment is docked at the node ; represents the set of all dockable nodes of mobile energy storage; based on the battery energy storage model established in step S1, the mobile energy storage operation constraints are established as follows:
[0176] ,
[0177] ,
[0178] ;
[0179] S28: The demand-side response mechanism applied to the node i load at time t, Represents the set of nodes that perform demand-side response; and are the load change and electricity price change before and after the demand side response is applied, is the electricity price variable after implementing demand response; is the electricity price elasticity coefficient, is the electricity price parameter before applying demand-side response, and are the upper and lower limits of the electricity price after applying demand-side response, Representation Node exist The active load at the moment; then the demand side response model is:
[0180] ,
[0181] ,
[0182] ,
[0183] ;
[0184] S29: Set It is the initial branch state of the distribution network system; It is the upper limit of branch switch action; is the total number of branch switches; then the switch constraint is:
[0185] .
[0186] S3: Use the historical data of renewable energy output to establish an ellipsoid uncertainty set, and establish a data-driven convex hull polyhedron set by connecting the vertices of the high-dimensional ellipsoid;
[0187] According to the constraints in step S2, there are uncertain variables ,set up represents the center point of the high-dimensional ellipsoid, is a positive definite matrix, which indicates the offset direction of the high-dimensional ellipsoid relative to the coordinate axis; the budget uncertainty set is used for the ellipsoid set The specific expression for describing the fluctuation range of distributed photovoltaic output is:
[0188] .
[0189] A data-driven convex hull uncertainty set is proposed. The specific steps are as follows:
[0190] S31: Construct a high-dimensional ellipsoid uncertainty set with the smallest volume and covering all historical data fluctuations ;
[0191] S32: Based on the original high-dimensional ellipsoid, the positive definite matrix Perform orthogonal decomposition and rotate and translate the original ellipsoid so that the center of the ellipsoid falls on the center of the coordinate axis. At this time, the high-dimensional ellipsoid uncertainty set is ;
[0192] S33: Scaling the high-dimensional linear polyhedron set in step S2 so that there is a minimum Make the scaled polyhedron set just envelop all possible data points, The value range is ;
[0193] set up Indicates The weight coefficient of each vertex; is the scaling factor, which is used to adjust the conservativeness of the envelope range of the high-dimensional linear polyhedron; then the uncertain set of the high-dimensional linear polyhedron after scaling is for:
[0194] ;
[0195] S34: Rotate and translate the scaled high-dimensional linear polyhedron to fit the original data point range; then the high-dimensional linear polyhedron after rotation and translation is an uncertain set for:
[0196] .
[0197] S4: The model established in step S2 is converted into a matrix form and the C&CG algorithm is used to solve the model.
[0198] The constraints in step S2 are written in matrix form as follows:
[0199] Power flow constraints: ;
[0200] Distributed photovoltaic operation constraints: ;
[0201] On-load tap-changing transformer constraints: ;
[0202] AC / DC converter station constraints: ;
[0203] Demand-side response constraints: ;
[0204] Battery storage constraints: ;
[0205] Constraints for switching capacitors in groups:
[0206] set up represents the decision variable in the first stage, represents the decision variable in the second stage, Indicates uncertain variables; , In the objective function , The relevant parameter matrix; , , , , , , , represents the parameter matrix related to the two-stage robust optimization constraints; Indicates a given When continuous variable The feasible domain of the first stage objective function represents the switching cost, the objective function of the second stage Represents the sum of the remaining costs except the switch cost; combined with the objective function, the matrix form of the dynamic robust reconstruction model of the two-stage AC / DC hybrid distribution network is established as follows:
[0207] .
[0208] The C&CG method is used to transform the dynamic robust reconfiguration model of two-stage AC / DC hybrid distribution network into a master-slave problem.
[0209] set up , , is the dual variable constrained by the subproblem, and the corresponding master-slave problem model is:
[0210] ,
[0211] ;
[0212] The outer approximation is used to solve the existing bilinear terms ( ) problem, according to the duality theorem, the inner layer minimization problem of SP1 is changed into the same maximization problem as the outer layer:
[0213] ;
[0214] set up is the number of historical iterations, is the current iteration number, As auxiliary variables, MP2 for solving the upper bound and SP2 for solving the lower bound are:
[0215] ,
[0216] ;
[0217] In the above formula The linearization formula is:
[0218] .
[0219] It should be understood that the size of the serial numbers of the steps in the above embodiments does not mean the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of the present application.
[0220] Example 2
[0221] In order to fully verify the feasibility and efficiency of the dynamic reconstruction method of the AC / DC hybrid distribution network proposed in Example 1, this example selects the improved IEEE-33 node system as an example for in-depth analysis. The system reference voltage is set to 12.66kV, and the reference capacity is set to 10MVA. Power range: active 0~2000kW, reactive 0~2000kvar. The current upper limit is 0.5pu, and the voltage range is 0.95~1.05pu.
[0222] In this embodiment, the IEEE-33 node system is divided into residential areas, industrial areas and commercial areas, and the demand-side response mechanism is only adopted for residential areas and commercial areas; at the initial moment, except for the tie line switch being disconnected, the other branch switches are all closed; the MESS is parked at the 5th node, and in order to facilitate subsequent analysis, the moving paths of the mobile energy storage are numbered: 1#~8#.
[0223] In order to verify the effectiveness of the mobile energy storage model under non-ideal conditions for the dynamic robust reconstruction method of the AC / DC hybrid distribution network, this embodiment sets three strategies for analysis:
[0224] Strategy 1: Use a battery energy storage model under non-ideal conditions and store energy in 1#.
[0225] Strategy 2: Use the traditional battery energy storage model, and the moving path is 1#~8#.
[0226] Strategy 3: Adopt a mobile energy storage model under non-ideal conditions, and the moving path is 1#~8#.
[0227] Table 1 gives the results of the three strategies in detail.
[0228] Table 1 Results of three strategies
[0229]
[0230] As can be seen from Table 1, in terms of network loss cost, the effectiveness of strategies 3 and 1 is significant, which is significantly better than strategy 1. This shows that compared with the traditional energy storage model, the battery energy storage model proposed in this paper helps to reduce network loss; and because strategy 1 is a fixed energy storage, it cannot accurately respond to complex power flow directions, so its network loss cost is slightly higher than strategy 3. In terms of abandoned light cost, the three strategies all use hyperplane uncertain sets, and the abandoned light cost is not much different; while strategy 1 uses fixed energy storage, which is similar to the network loss cost. Due to the significant spatial performance of mobile energy storage and its strong absorption capacity for uncertain energy, the abandoned light cost of strategies 2 and 3 is slightly lower than that of strategy 1. In terms of power purchase cost, strategy 3 < strategy 1 < strategy 2, which shows that the use of mobile energy storage under non-ideal conditions helps to optimize the purchase of power at the gateway, and can greatly reduce the purchase of electricity while reducing network loss. For the switch cost, the number of distribution network switch actions depends on the power flow changes of the system, and the addition of mobile energy storage will balance the power flow well, so the switch cost of strategies 3 and 2 is lower than that of strategy 1. From the perspective of total cost, the total system cost of Strategy 3 is significantly better than that of Strategy 1 and Strategy 2. This phenomenon strongly proves that the mobile energy storage model proposed by the present invention and the battery energy storage model under non-ideal conditions have significant advantages in economic benefits and practicality. Furthermore, the latter performs even better in terms of economic efficiency.
[0231] The above embodiments are only used to illustrate the design ideas and features of the present invention, and their purpose is to enable those skilled in the art to understand the content of the present invention and implement it accordingly. The protection scope of the present invention is not limited to the above embodiments. Therefore, any equivalent changes or modifications made based on the principles and design ideas disclosed by the present invention are within the protection scope of the present invention.
Claims
1. A method for dynamic robust reconstruction of distribution network based on mobile energy storage under non-ideal conditions, characterized by: The following steps are involved: S1: Establish the relationship between the upper and lower limits of charge and discharge power and the charge and discharge current, the input power of the battery and the output power of the battery from the charge and discharge process inside the battery; sample the operation area of the charge and discharge power, and transform the reconstructed battery energy storage model into a linear model in the sampling point set of the convex envelope; For nodes Department The battery energy storage connected during the time period is set is the input power of the battery, is the output power of the battery; and is a variable associated with the charge and discharge sample data; for the discharge curve, let , , They represent the battery energy storage discharge sample sampling, J is the sampling point set of the discharge sample data; for the charging curve, let , , They represent the battery energy storage charging sample sampling, and K is the sampling point set of the charging sample data; the battery energy storage model under non-ideal conditions is expressed as a linear combination of sample sampling data: , , , , , , , ; S2: Select the minimization of the comprehensive cost within the dispatching period as the objective function and establish a dynamic reconstruction model of the AC / DC hybrid distribution network; The network loss cost coefficient is The switching operation cost coefficient is expressed as The cost coefficient of abandoned light is expressed as The power purchase cost coefficient is expressed as express; is a Boolean variable representing Branch circuit in the distribution network system The breaking condition of the branch ,definition Time from the beginning Flow to the end The power flow is , from the end Flow to the head end The power flow is ; , is with The physical quantities related to the output of the distributed photovoltaic model at each moment are the available power and the actual output power after the distributed photovoltaic fluctuations; for The power purchased at the time limit; represents the comprehensive cost function of distribution network reconstruction; Represents the set of all branches of the distribution network; represents the time set of the distribution network reconstruction; Represents the node set of distributed photovoltaic access; for the dynamic reconstruction model of AC / DC hybrid distribution network, the minimization of the comprehensive cost within the dispatching period is selected as the objective function: , The first term in the above formula represents the system network loss cost, the second term represents the system switch action cost, the third term represents the system abandoned light cost, and the fourth term represents the battery energy storage aging cost; The constraints include power flow constraints, on-load tap-changing transformer constraints, radial topology, AC / DC converter station constraints, capacitor constraints, distributed photovoltaic operation constraints, mobile energy storage operation constraints, demand-side response model and switch constraints. The specific calculation steps are as follows: S21: Pair Node exist The power flow variables at time are The total active power injected into the node, The total reactive power injected into the node, Power injected into mobile energy storage, Active load after applying demand-side response mechanism to the node, is the reactive load of the node, The reactive power injected into the gateway, is the reactive power generated by the capacitor; definition Time from the beginning Flow to the end The reactive power is , from the end Flow to the head end The reactive power is ; For all nodes A collection of connected nodes; for branches ,set up is the branch current amplitude, , is the physical parameter of the branch; definition , , ,in, , is the voltage across the branch ij, is the voltage phase difference at both ends, and the variable The variables related to the branch switch state and the branch terminal voltage are expressed as , and , The relationship between them; then the power flow constraint is: , , , , , , , , , ; Set Node The maximum voltage at , the minimum value is , add additional constraints as follows: , ; Set up branch road The maximum current is , the branch current constraint is: ; S22: Set is the ratio parameter, is the gear number parameter, is the gear interval value parameter, is the upper limit parameter of the ratio, is the lower limit parameter of the transformation ratio; is an integer variable representing the transformer gear position; , , ; Then the voltage relationship between the two ends of the transformer, that is, the on-load tap-changing transformer constraint is: , , , ; S23: Set Represents a set of distribution network nodes. Represents the root node set; , Both are Boolean variables used to judge nodes and nodes The relationship between them; then the radial topology is: ; S24: In the AC / DC converter station model, , They are The power flow measured by the AC converter station at every moment; for The DC power flow is measured at the converter station at all times; is the reference point voltage, let , ; is a set of converter branches; Indicates the DC side voltage utilization parameter; represents the modulation parameter; represents the conversion rate parameter; then the power flow equation of the AC / DC converter station model, i.e., the AC / DC converter station constraint, is: , , , , , , , ; S25: For nodes exist The capacitor bank connected at the moment is set A node assembly for installing capacitors; is the final number of grid-connected capacitor banks; The upper limit of the number of capacitor groups is is the capacity of a single group of capacitors, is the maximum value of the change in the number of capacitor groups during the interval period; then the capacitor constraint is: , ; S26: For nodes exist Distributed photovoltaic connected at all times, is the maximum available power of distributed photovoltaics before fluctuation; is the maximum fluctuation of distributed photovoltaic, For distributed photovoltaic Uncertain variables at the moment; For distributed photovoltaic Reactive power at the moment, is the capacity of distributed photovoltaics; then the distributed photovoltaic operation constraints are: , , ; S27: Set It is a Boolean variable, indicating that the mobile storage Whether the moment is docked at the node ; represents the set of all dockable nodes of mobile energy storage; based on the battery energy storage model established in step S1, the mobile energy storage operation constraints are established as follows: , , ; S28: The demand-side response mechanism applied to the node i load at time t, Represents the set of nodes that perform demand-side response; and are the load change and electricity price change before and after the demand side response is applied, is the electricity price variable after implementing demand response; is the electricity price elasticity coefficient, is the electricity price parameter before applying demand-side response, and are the upper and lower limits of the electricity price after applying demand-side response, Representation Node exist The active load at the moment; then the demand side response model is: , , , ; S29: Set It is the initial branch state of the distribution network system; It is the upper limit of branch switch action; is the total number of branch switches; then the switch constraint is: ; S3: Use the historical data of renewable energy output to establish an ellipsoid uncertainty set, and establish a data-driven convex hull polyhedron set by connecting the vertices of the high-dimensional ellipsoid; According to the constraints in step S2, there are uncertain variables ,set up represents the center point of the high-dimensional ellipsoid, is a positive definite matrix, which indicates the offset direction of the high-dimensional ellipsoid relative to the coordinate axis; the budget uncertainty set is used for the ellipsoid set The specific expression for describing the fluctuation range of distributed photovoltaic output is: ; A data-driven convex hull uncertainty set is proposed. The specific steps are as follows: S31: Construct a high-dimensional ellipsoid uncertainty set with the smallest volume and covering all historical data fluctuations ; S32: Based on the original high-dimensional ellipsoid, the positive definite matrix Perform orthogonal decomposition and rotate and translate the original ellipsoid so that the center of the ellipsoid falls on the center of the coordinate axis. At this time, the high-dimensional ellipsoid uncertainty set is ; S33: Scaling the high-dimensional linear polyhedron set in step S32 so that there is a minimum Make the scaled polyhedron set just envelop all possible data points, The value range is ; set up Indicates The weight coefficient of each vertex; is the scaling factor, which is used to adjust the conservativeness of the envelope range of the high-dimensional linear polyhedron; then the uncertain set of the high-dimensional linear polyhedron after scaling is for: ; S34: Rotate and translate the scaled high-dimensional linear polyhedron to fit the original data point range; then the high-dimensional linear polyhedron after rotation and translation is an uncertain set for: ; S4: transforming the model established in step S2 in matrix form, covering historical data based on the convex hull polyhedron set in step S3, and solving the model using the C&CG algorithm; The constraints in step S2 are written in matrix form as follows: Power flow constraints: ; Distributed photovoltaic operation constraints: ; On-load tap-changing transformer constraints: ; AC / DC converter station constraints: ; Demand-side response constraints: ; Battery storage constraints: ; Constraints for switching capacitors in groups: set up represents the decision variable in the first stage, represents the decision variable in the second stage, Indicates uncertain variables; , In the objective function , The relevant parameter matrix; , , , , , , , represents the parameter matrix related to the two-stage robust optimization constraints; Indicates a given When continuous variable The feasible domain of the first stage objective function represents the switching cost, the objective function of the second stage Represents the sum of the remaining costs except the switch cost; combined with the objective function, the matrix form of the dynamic robust reconstruction model of the two-stage AC / DC hybrid distribution network is established as follows: ; The C&CG method is used to transform the dynamic robust reconfiguration model of two-stage AC / DC hybrid distribution network into a master-slave problem. set up , , is the dual variable constrained by the subproblem, and the corresponding master-slave problem model is: , ; The outer approximation is used to solve the existing bilinear terms ( ) problem, according to the duality theorem, the inner minimization problem of SP1 is changed into the same maximization problem as the outer layer: ; set up is the number of historical iterations, is the current iteration number, As auxiliary variables, MP2 for solving the upper bound and SP2 for solving the lower bound are: , ; In the above formula The linearization formula is: 。 2. A dynamic robust reconstruction system for a distribution network based on mobile energy storage under non-ideal conditions, the system comprising a processor and a memory, characterized in that: The memory stores computer instructions, and the processor is used to execute the computer instructions stored in the memory. When the computer instructions are executed by the processor, the system implements the steps of the method for dynamic robust reconstruction of a distribution network based on mobile energy storage under non-ideal conditions as described in claim 1.
3. A computer memory, characterized in that: A computer program executable by a computer processor is stored therein, and the computer program executes the method for dynamic robust reconstruction of a distribution network based on mobile energy storage under non-ideal conditions as described in claim 1.
