A segmented multi-level cloud-edge coordinated operation control method for distribution network

Abstract

The present invention discloses a method for controlling the coordinated operation of a segmented multi-level cloud-edge in a distribution network, comprising the following steps: based on the rapid identification result of the reconstruction level, a time-divided multi-level reconstruction model is established with the goal of minimizing the cost of abandoned light, lost load, and switch action, and the time-divided multi-level reconstruction model is converted into a mixed integer second-order cone programming model through second-order cone relaxation; a typical scenario set is used to characterize the uncertainty fluctuations of photovoltaics and loads, and the mixed integer second-order cone programming model is optimized through controllable loads to obtain a coordinated optimization model based on the scenario set; a centralized-distributed joint control and calculation system of cloud and edge nodes is established, and the coordinated operation control of the segmented multi-level cloud-edge in the distribution network is realized through a coordinated optimization model based on a scenario set. The present invention can significantly reduce the data communication pressure and computing energy consumption of the traditional reconstruction method of the distribution network.

Description

Technical Field

[0001] The present invention relates to the field of distribution network operation optimization, and in particular to a distribution network segmented multi-level cloud-edge coordinated operation control method. Background Art

[0002] The centralized optimization method adopted by global dynamic reconstruction requires that the collected terminal data be uniformly uploaded to the distribution network dispatching center for unified decision-making on the switch action plan. At the current stage, with the increasing scale and network complexity of the distribution network, it has problems such as high data communication pressure, high computing energy consumption, and difficulty in model solution. The segmented multi-level reconstruction method that prefers local autonomy implements hierarchical local reconstruction with transformers, substations, and multi-substations as reconstruction objects, which can effectively reduce the difficulty of model solution while increasing clean energy consumption and ensuring reliable power supply of loads. Multi-level reconstruction realizes fully distributed optimization, and the dispatching center does not need to calculate the whole network reconstruction model. It can be seen that the data communication pressure, computing energy consumption and model solution difficulty of the whole network centralized optimization reconstruction can be significantly reduced through the joint optimization of centralized reconstruction level identification and distributed multi-level reconstruction. Therefore, it is of great significance to construct a cloud-edge coordinated operation control architecture under the segmented multi-level reconstruction of the distribution network. The centralized-distributed joint control and computing system of cloud-edge collaboration can provide a feasible path for the implementation of segmented multi-level reconstruction, and establish edge computing groups at each layer of the distribution network to realize hierarchical autonomous collaboration. Summary of the invention

[0003] The purpose of the present invention is to overcome the shortcomings of the prior art and provide a distribution network segmented multi-level cloud-edge coordinated operation control method, comprising the following steps:

[0004] Based on the rapid identification results of the reconstruction level, a time-division multi-level reconstruction model is established with the goal of minimizing the cost of abandoned light, lost load, and switch action. The time-division multi-level reconstruction model is then transformed into a mixed integer second-order cone programming model through second-order cone relaxation.

[0005] A typical scenario set is used to characterize the uncertainty fluctuations of photovoltaic power generation and load. The mixed integer second-order cone programming model is optimized through controllable loads to obtain a coordinated optimization model based on the scenario set.

[0006] A centralized-distributed joint control and computing system for cloud and edge nodes is established, and segmented multi-level cloud-edge coordinated operation control of the distribution network is achieved through a coordinated optimization model based on a scenario set.

[0007] Furthermore, the establishment of a time-divided multi-level reconstruction model with the goal of minimizing the cost of abandoning light, losing load, and switching action includes the following processes: substation-level, transformer-level, and feeder-level reconstruction models with the goal of minimizing the cost of abandoning light, losing load, and switching action, including the reconstruction level being identified as the substation level: taking multiple substations as objects, realizing multi-station net load power supply through the state combination of feeder tie switches, transformer tie switches, substation tie switches, and section switches;

[0008] The reconstruction level is identified as transformer level: taking the substation as the object, the autonomous operation of the substation unit is realized through the state combination of feeder tie switches, transformer tie switches and section switches;

[0009] The reconstruction level is identified as the feeder level: taking the transformer as the object, the power supply of the feeder load in the transformer and the full absorption of photovoltaic power are realized through the state combination of the feeder tie switch and the section switch.

[0010] Furthermore, the objective function of the substation-level reconstruction model of the time-division multi-level reconstruction model is:

[0011]

[0012]

[0013]

[0014]

[0015] Where: represents the objective function of substation-level reconstruction of substations m and n in the first stage of the c-th period, T c represents the set of time periods belonging to the cth category; C PV , C LR , C SW They represent the cost of abandoned light, the cost of load reduction, and the cost of switch reconfiguration respectively; are the single operation costs of feeder tie switches, transformer tie switches, inter-station tie switches, and section switches, respectively, and Indicates the switch state change flag of branch ij during period t. This means that the switch of branch ij changes from open state to closed state during period t; if This means that the switch of branch ij changes from closed state to open state during period t; E SW,Feed 、E SW,Trans 、E SW,Sub They represent the sets of various types of tie switch branches in the reconstruction area, which are adjusted according to different substation-level reconstructions; B, B PV They represent the node set of the reconstruction area and the PV node set respectively.

[0016] Furthermore, the objective function of the transformer-level reconstruction model of the segmented multi-level reconstruction model is:

[0017]

[0018]

[0019] Where: It represents the objective function of transformer-level reconstruction of substation j in the first stage in the cth period.

[0020] Furthermore, the coordination optimization model based on the scenario set is:

[0021] Monte Carlo simulation method and K-means algorithm are used to generate typical scenario sets to express the uncertainty output of photovoltaic and load. The dynamic coordination optimization model based on the scenario set can be represented in matrix form as follows:

[0022]

[0023] stAx≤b

[0024] Cx=d

[0025] Ex+Fy s =g

[0026] ||Hy s ||2≤k Τ y s

[0027] R s ≤ξ s

[0028] Jy s ≤l

[0029] Oy s =q

[0030] Where: s is the scene flag, N s is the total number of scenes, f s is the scenario probability; x is the discrete decision variable, including the controllable load related variables; y s It is a real-time adjustable variable, including state variables, substation output, abandoned light, and load loss related variables; s is the active value of photovoltaic power generation and load demand in scenario s; Ax≤b and Cx=d ​​correspond to controllable load constraints; Ex+Fy s =g corresponds to the power flow balance equation constraint formula of the associated discrete variables and real-time decision variables; formula || Hy s ||2≤k Τ y sRepresents the second-order cone constraint of the tidal current; Formula Ry s ≤ξ s represents PV constraint and load loss constraint; formula Jy s ≤l,Oy s =q represent the remaining equality constraints and inequality constraints involving real-time adjustable variables.

[0031] Furthermore, the centralized-distributed joint control and computing system for cloud and edge nodes is established, and the segmented multi-level cloud-edge coordinated operation control of the distribution network is realized through a coordinated optimization model based on a scenario set, including the following process:

[0032] The centralized-distributed joint control computing system of the cloud and edge nodes is that the edge computing control platform uploads the net load prediction data of each node to the big data center of the cloud platform through the backbone communication network. The cloud completes the reconstruction level identification, and converts the reconstruction level requirements of each substation and transformer into a centralized control command, which is sent to each edge computing control platform. The reconstruction of each segmented feeder level and transformer level is executed by the corresponding substation edge computing control platform, and the switch action plan is converted into a control command to remotely control each switch. The substation-level reconstruction is executed by the cloud platform, and the switch decision results are converted into centralized control commands to be executed separately by the relevant substation edge nodes.

[0033] The beneficial effects of the present invention are: the present invention explores a segmented multi-level cloud-edge coordinated operation control method for a distribution network, establishes a segmented multi-level reconstruction model based on the reconstruction level identification results of each substation and transformer, and compensates for the deficiency of the segmented multi-level reconstruction flow distribution improvement capability through dynamic coordinated optimization of controllable loads, and finally realizes centralized reconstruction level identification and distributed multi-level reconstruction based on a centralized-distributed joint optimization framework of cloud-edge collaboration, which significantly reduces the data communication pressure, computational energy consumption and model solving difficulty of traditional reconstruction methods of distribution networks. BRIEF DESCRIPTION OF THE DRAWINGS

[0034] Figure 1 It is a schematic diagram of the flow chart of the segmented multi-level cloud-edge coordinated operation control method of the present invention;

[0035] Figure 2 It is a segmented multi-level reconstruction flow chart of the present invention;

[0036] Figure 3 This is a control framework diagram of the cloud-edge coordinated operation of the distribution network of the present invention;

[0037] Figure 4 It is the cloud-based reconstruction level recognition result of the present invention;

[0038] Figure 5 It is the cloud-based segmented multi-level reconstruction level recognition result of the present invention. DETAILED DESCRIPTION

[0039] The technical solution of the present invention is further described in detail below in conjunction with the accompanying drawings, but the protection scope of the present invention is not limited to the following.

[0040] like Figure 1 As shown, a distribution network segmented multi-level cloud-edge coordinated operation control method includes the following steps:

[0041] Based on the rapid identification results of the reconstruction level, a time-division multi-level reconstruction model is established with the goal of minimizing the cost of abandoned light, lost load, and switch action. The time-division multi-level reconstruction model is then transformed into a mixed integer second-order cone programming model through second-order cone relaxation.

[0042] A typical scenario set is used to characterize the uncertainty fluctuations of photovoltaic power generation and load. The mixed integer second-order cone programming model is optimized through controllable loads to obtain a coordinated optimization model based on the scenario set.

[0043] A centralized-distributed joint control and computing system for cloud and edge nodes is established, and segmented multi-level cloud-edge coordinated operation control of the distribution network is achieved through a coordinated optimization model based on a scenario set.

[0044] The multi-level reconstruction model for different time periods is established with the goal of minimizing the cost of abandoning light, losing load and switching action, including the following processes: substation-level, transformer-level and feeder-level reconstruction models with the goal of minimizing the cost of abandoning light, losing load and switching action, including the reconstruction level identification as substation level: taking multiple substations as objects, realizing the power supply of multiple station net loads through the state combination of feeder tie switches, transformer tie switches, substation tie switches and section switches;

[0045] The reconstruction level is identified as transformer level: taking the substation as the object, the autonomous operation of the substation unit is realized through the state combination of feeder tie switches, transformer tie switches and section switches;

[0046] The reconstruction level is identified as the feeder level: taking the transformer as the object, the power supply of the feeder load in the transformer and the full absorption of photovoltaic power are realized through the state combination of the feeder tie switch and the section switch.

[0047] The objective function of the substation-level reconstruction model of the time-division multi-level reconstruction model is:

[0048]

[0049]

[0050]

[0051]

[0052] Where: represents the objective function of substation-level reconstruction of substations m and n in the first stage of the c-th period, T c represents the set of time periods belonging to the cth category; C PV , C LR , C SW They represent the cost of abandoned light, the cost of load reduction, and the cost of switch reconfiguration respectively; are the single operation costs of feeder tie switches, transformer tie switches, inter-station tie switches, and section switches, respectively, and Indicates the switch state change flag of branch ij during period t. This means that the switch of branch ij changes from open state to closed state during period t; if This means that the switch of branch ij changes from closed state to open state during period t; E SW,Feed 、E SW,Trans 、E SW,Sub They represent the sets of various types of tie switch branches in the reconstruction area, which are adjusted according to different substation-level reconstructions; B, B PV They represent the node set of the reconstruction area and the PV node set respectively.

[0053] The objective function of the transformer-level reconstruction model of the segmented multi-level reconstruction model is:

[0054]

[0055]

[0056] Where: It represents the objective function of transformer-level reconstruction of substation j in the first stage in the cth period.

[0057] The coordination optimization model based on the scenario set is:

[0058] Monte Carlo simulation method and K-means algorithm are used to generate typical scenario sets to express the uncertainty output of photovoltaic and load. The dynamic coordination optimization model based on the scenario set can be represented in matrix form as follows:

[0059]

[0060] stAx≤b

[0061] Cx=d

[0062] Ex+Fy s =g

[0063] ||Hy s ||2≤k Τ y s

[0064] R s ≤ξ s

[0065] Jy s ≤l

[0066] Oy s =q

[0067] Where: s is the scene flag, N s is the total number of scenes, f s is the scenario probability; x is the discrete decision variable, including the controllable load related variables; y s It is a real-time adjustable variable, including state variables, substation output, abandoned light, and load loss related variables; s is the active value of photovoltaic power generation and load demand in scenario s; Ax≤b and Cx=d ​​correspond to controllable load constraints; Ex+Fy s =g corresponds to the power flow balance equation constraint formula of the associated discrete variables and real-time decision variables; formula || Hy s ||2≤k Τ y s Represents the second-order cone constraint of the tidal current; Ry s ≤ξ s represents PV constraint and load loss constraint; formula Jy s ≤l,Oy s =q represent the remaining equality constraints and inequality constraints involving real-time adjustable variables.

[0068] The centralized-distributed joint control and computing system for establishing cloud and edge nodes is described, and the segmented multi-level cloud-edge coordinated operation control of the distribution network is realized through a coordinated optimization model based on a scenario set, including the following processes:

[0069] The centralized-distributed joint control computing system of the cloud and edge nodes is that the edge computing control platform uploads the net load prediction data of each node to the big data center of the cloud platform through the backbone communication network. The cloud completes the reconstruction level identification, and converts the reconstruction level requirements of each substation and transformer into a centralized control command, which is sent to each edge computing control platform. The reconstruction of each segmented feeder level and transformer level is executed by the corresponding substation edge computing control platform, and the switch action plan is converted into a control command to remotely control each switch. The substation-level reconstruction is executed by the cloud platform, and the switch decision results are converted into centralized control commands to be executed separately by the relevant substation edge nodes.

[0070] Specifically, the segmented multi-level cloud-edge coordinated operation control method for the distribution network is based on the rapid identification results of the reconstruction level, and a segmented multi-level reconstruction model is established with the goal of minimizing the cost of abandoned light, loss of load, and switch action, and is converted into a mixed integer second-order cone programming model through second-order cone relaxation; a typical scenario set is used to characterize the uncertainty fluctuations of photovoltaics and loads, and dynamic coordination and optimization of controllable loads are performed through active management measures; a new cloud-edge coordinated control method is proposed, which serves the operation optimization control of the segmented multi-level reconstructed distribution network through a centralized-distributed joint control computing system; the effectiveness of the segmented multi-level cloud-edge coordinated operation control method proposed in the present invention is verified based on a calculation example system.

[0071] The segmented multi-level reconstruction model:

[0072] like Figure 2 As shown, segmented multi-level reconstruction is performed with the goal of minimizing the cost of abandoned light, lost load, and switching action, and the full-time switching action plan is retained.

[0073] The substation level reconstruction model:

[0074] The objective function is:

[0075]

[0076]

[0077]

[0078]

[0079] Where: represents the objective function of substation-level reconstruction of substations m and n in the first stage of the c-th period, T c represents the set of time periods belonging to the cth category; C PV , C LR , C SW They represent the cost of abandoned light, the cost of load reduction, and the cost of switch reconfiguration respectively; are the single operation costs of feeder tie switches, transformer tie switches, inter-station tie switches, and section switches, respectively, and Indicates the switch state change flag of branch ij during period t. This means that the switch of branch ij changes from open state to closed state during period t; if This means that the switch of branch ij changes from closed state to open state during period t; E SW,Feed 、E SW,Trans 、E SW,Sub They represent the sets of various types of tie switch branches in the reconstruction area, which are adjusted according to different substation-level reconstructions; B, B PVThey represent the node set of the reconstruction area and the PV node set respectively.

[0080] The constraints are:

[0081] (1) Second-order cone power flow constraint

[0082] The original nonlinear power flow constraint is transformed into a second-order cone power flow constraint through second-order cone relaxation:

[0083]

[0084]

[0085]

[0086]

[0087]

[0088]

[0089]

[0090]

[0091]

[0092] Where: α(j) is the set of terminal nodes of the branch with j as the initial node; β(j) is the set of initial nodes of the branch with j as the terminal node; r ij 、x ij and g j , b j Respectively represent branch impedance and node admittance; V j,t is the node voltage; They represent the active load and load reduction active power of node j in the tth period respectively; They represent the reactive power injection, reactive load and load reduction of the substation at node j in the tth period respectively; P ij,t , Q ij,t They represent the active and reactive power of branch ij in the tth time period respectively.

[0093] (2) Safety constraints

[0094]

[0095]

[0096]

[0097] Where: and Iij are the upper and lower limits of the current of branch ij; and V j are the upper and lower limits of the voltage at node j respectively.

[0098] (3) Network Reconfiguration Constraints

[0099]

[0100]

[0101]

[0102] For N Sub For a B-node system with 100 substations, the number of closed branches in the network is the number of nodes minus the number of substations. At the same time, a smaller node injection power is introduced at non-substation nodes, and the network connectivity is guaranteed by adding auxiliary power flow constraints.

[0103] (4) Substation power supply constraints

[0104] Since each transformer in the substation is powered independently, the substation power constraint is expressed as the physical property constraint of the transformer.

[0105]

[0106]

[0107] Where: B Sub Represents the set of substation nodes in the reconstruction area. In addition, PV output constraints and load loss constraints are not described here.

[0108] The transformer-level reconstruction model:

[0109] The objective function:

[0110]

[0111]

[0112] Where: It represents the objective function of transformer-level reconstruction of substation j in the c-th time period in the first stage.

[0113] The constraints are:

[0114] The constraints are similar to those of substation-level reconstruction, including power flow constraints, safety constraints, network reconstruction constraints, PV constraints, substation power supply constraints, and load loss constraints. The network reconstruction constraints and substation power supply constraints are different from the above and are changed to the following formula:

[0115]

[0116]

[0117]

[0118] The feeder level reconstruction model:

[0119] The objective function:

[0120]

[0121]

[0122] Where: It represents the objective function of feeder-level reconstruction of transformer f in substation j during the c-th period in the first stage.

[0123] The constraints are:

[0124] The constraints are similar to those of substation-level reconstruction, including power flow constraints, safety constraints, network reconstruction constraints, PV constraints, substation power supply constraints, and load loss constraints. The network reconstruction constraints and substation power supply constraints are different from the above and are changed to the following formula:

[0125]

[0126]

[0127]

[0128] The coordinated optimization model based on the scenario set:

[0129] Monte Carlo simulation method and K-means algorithm are used to generate typical scenario sets to express the uncertainty output of photovoltaic and load. The dynamic coordination optimization model based on the scenario set can be represented in matrix form as follows:

[0130]

[0131] stAx≤b

[0132] Cx=d

[0133] Ex+Fy s =g

[0134] ||Hy s ||2≤k Τ y s

[0135] R s ≤ξ s

[0136] Jys ≤l

[0137] Oy s =q

[0138] Where: s is the scene flag, N s is the total number of scenes, f s is the scenario probability; x is the discrete decision variable, including the controllable load related variables; y s It is a real-time adjustable variable, including state variables, substation output, abandoned light, and load loss related variables; s is the active value of photovoltaic power generation and load demand in scenario s; Ax≤b and Cx=d ​​correspond to controllable load constraints; Ex+Fy s =g corresponds to the power flow balance equation constraint formula of the associated discrete variables and real-time decision variables; formula || Hy s ||2≤k Τ y s Represents the second-order cone constraint of the tidal current; Formula Ry s ≤ξ s represents PV constraint and load loss constraint; formula Jy s ≤l,Oy s =q represent the remaining equality constraints and inequality constraints involving real-time adjustable variables.

[0139] The controllable load constraints:

[0140]

[0141]

[0142]

[0143]

[0144]

[0145]

[0146] Where: They respectively represent the upper limits of the positive and negative regulation of the controllable load at node j within the total time period.

[0147] The cloud-edge coordinated operation control method:

[0148] The regional power distribution system powered by substations will serve as the basic unit, and all physical and information devices in the area will form a distributed edge service platform that integrates network, computing, storage, and application functions. That is, edge computing servers are configured at each substation site to share the computing tasks of the cloud computing center. Figure 3As shown in the figure, the collaborative working mechanism of the cloud and edge nodes is as follows: 1) The edge computing control platform uploads the net load prediction data of each node to the big data center of the cloud platform through the backbone communication network. The cloud platform completes the reconstruction level identification, converts the reconstruction level requirements of each substation and transformer into centralized control commands, and sends them to each edge computing control platform. 2) The reconstruction of each segmented feeder level and transformer level is executed by the corresponding substation edge computing control platform, and the switch action plan is converted into a control command to remotely control each switch. The substation-level reconstruction is executed by the cloud platform, and the switch decision results are converted into centralized control commands and executed separately by the relevant substation edge nodes. It can be seen that the cloud-edge collaborative mode can not only meet the centralized optimization requirements of level identification and substation-level reconstruction, but also is suitable for the decentralized autonomy of segmented multi-level reconstruction.

[0149] The example verification analysis:

[0150] Based on the improved 445-node system, the segmented multi-level cloud-edge coordinated operation control method of the distribution network was verified and analyzed on four laptops and one desktop computer. Each server was configured with a MATLAB 2016a simulation platform with built-in YALMIP toolbox and CPLEX algorithm package.

[0151] Cloud reconstruction level recognition results:

[0152] like Figure 4 As shown in the figure, substations S1 to S4 do not need substation-level reconstruction in all time periods. Transformer-level reconstruction and feeder-level reconstruction can effectively alleviate the phenomenon of abandoned light and load loss caused by the uneven temporal and spatial distribution of net load. It is worth noting that the photovoltaic power generation in the 9th period cannot be further absorbed by substation-level reconstruction. The reconstruction level identification effectively avoids the low progressive benefits brought by high-level reconstruction. The level identification results after the reconstruction period division are shown in Figure 1. Figure 5 As shown, the difficulty of implementing segmented multi-level cloud-edge coordinated operation control is further reduced.

[0153] Segmented multi-level reconstruction:

[0154] According to the cloud-level recognition results, segmented multi-level reconstruction is performed, and the full-time switch action plan calculated by the edge nodes deployed in each substation is shown in Table 1. Under segmented multi-level reconstruction, the amount of abandoned light and load loss is reduced from the original 20.8666MWh and 11.1744MWh to 0.2297MWh and 0MWh.

[0155] Table 1 Switch action scheme

[0156]

[0157]

[0158] Comparison of solution time of different reconstruction methods:

[0159] The solution time of different reconstruction methods applied to large-scale distribution network operation control is shown in Table 2. It can be seen that the proposed distribution network segmented multi-level cloud-edge coordinated operation control method has the shortest solution time, which can significantly reduce the difficulty of solving mixed integer nonlinear models and avoid the dimensionality disaster and high data transmission bandwidth caused by global dynamic reconstruction. Compared with the centralized segmented multi-level reconstruction method, cloud-edge collaboration realizes true decentralized computing through computing task allocation. It is worth noting that the proposed cloud-edge coordinated operation control method will show greater advantages when applied to larger-scale distribution networks.

[0160] Table 2 Comparison of solution time of different reconstruction methods

[0161]

[0162] The above is only a preferred embodiment of the present invention. It should be understood that the present invention is not limited to the form disclosed herein, and should not be regarded as excluding other embodiments, but can be used in various other combinations, modifications and environments, and can be modified within the scope of the concept described herein through the above teachings or the technology or knowledge of the relevant field. The changes and modifications made by those skilled in the art shall not deviate from the spirit and scope of the present invention, and shall be within the scope of protection of the claims attached to the present invention.

Claims

1. A distribution network segmented multi-level cloud-edge coordinated operation control method, characterized in that: The steps include: Based on the rapid identification results of the reconstruction level, a time-division multi-level reconstruction model is established with the goal of minimizing the cost of abandoned light, lost load, and switch action. The time-division multi-level reconstruction model is then transformed into a mixed integer second-order cone programming model through second-order cone relaxation. A typical scenario set is used to characterize the uncertainty fluctuations of photovoltaic power generation and load. The mixed integer second-order cone programming model is optimized through controllable loads to obtain a coordinated optimization model based on the scenario set. Establish a centralized-distributed joint control and computing system for cloud and edge nodes, and realize segmented multi-level cloud-edge coordinated operation control of the distribution network through a coordinated optimization model based on scenario sets; The objective function of the substation-level reconstruction model of the time-division multi-level reconstruction model is: Where: represents the objective function of substation-level reconstruction of substations m and n in the first stage of the c-th period, T c represents the set of time periods belonging to the cth category; C PV , C LR , C SW They represent the cost of abandoned light, the cost of load reduction, and the cost of switch reconfiguration respectively; are the single operation costs of feeder tie switches, transformer tie switches, inter-station tie switches, and section switches, respectively, and Indicates the switch state change flag of branch ij during period t. This means that the switch of branch ij changes from open state to closed state during period t; if This means that the switch of branch ij changes from closed state to open state during period t; E SW,Feed 、E SW,Trans 、E SW,Sub They represent the sets of various types of tie switch branches in the reconstruction area, which are adjusted according to different substation-level reconstructions; B, B PV They represent the node set and PV node set of the reconstruction area respectively; The objective function of the transformer-level reconstruction model of the segmented multi-level reconstruction model is: Where: It represents the objective function of transformer-level reconstruction of substation j in the first stage in the c-th period; The objective function of the feeder-level reconstruction model of the segmented multi-level reconstruction model is: Where: It represents the objective function of feeder-level reconstruction of transformer f in substation j during the c-th period in the first stage.

2. A distribution network segmented multi-level cloud-edge coordinated operation control method according to claim 1, characterized in that: The multi-level reconstruction model for different time periods is established with the goal of minimizing the cost of abandoning light, losing load and switching action, including the following processes: substation-level, transformer-level and feeder-level reconstruction models with the goal of minimizing the cost of abandoning light, losing load and switching action, including the reconstruction level identification as substation level: taking multiple substations as objects, realizing the power supply of multiple station net loads through the state combination of feeder tie switches, transformer tie switches, substation tie switches and section switches; The reconstruction level is identified as transformer level: taking the substation as the object, the autonomous operation of the substation unit is realized through the state combination of feeder tie switches, transformer tie switches and section switches; The reconstruction level is identified as the feeder level: taking the transformer as the object, the power supply of the feeder load in the transformer and the full absorption of photovoltaic power are realized through the state combination of the feeder tie switch and the section switch.

3. A distribution network segmented multi-level cloud-edge coordinated operation control method according to claim 1, characterized in that: The coordination optimization model based on the scenario set is: Monte Carlo simulation method and K-means algorithm are used to generate typical scenario sets to express the uncertainty output of photovoltaic and load. The dynamic coordination optimization model based on the scenario set can be represented in matrix form as follows: stAx≤b Cx=d Ex+Fy s =g ||Hy s ||2≤k Τ the s Ry s ≤ξ s You s ≤l Moon s =q Where: s is the scene flag, N s is the total number of scenes, f s is the scenario probability; x is a discrete decision variable, including controllable load-related variables; y s It is a real-time adjustable variable, including state variables, substation output, abandoned light, and load loss related variables; s is the active value of photovoltaic power generation and load demand in scenario s; Ax≤b and Cx=d ​​correspond to controllable load constraints; Ex+Fy s =g corresponds to the power flow balance equation constraint formula of the associated discrete variables and real-time decision variables; formula || Hy s ||2≤k Τ y s Represents the second-order cone constraint of the tidal current; Ry s ≤ξ s represents PV constraint and load loss constraint; formula Jy s ≤l,Oy s =q represent the remaining equality constraints and inequality constraints involving real-time adjustable variables.

4. A distribution network segmented multi-level cloud-edge coordinated operation control method according to claim 1, characterized in that: The centralized-distributed joint control and computing system for establishing cloud and edge nodes is described, and the segmented multi-level cloud-edge coordinated operation control of the distribution network is realized through a coordinated optimization model based on a scenario set, including the following processes: The centralized-distributed joint control computing system of the cloud and edge nodes is that the edge computing control platform uploads the net load prediction data of each node to the big data center of the cloud platform through the backbone communication network. The cloud completes the reconstruction level identification, and converts the reconstruction level requirements of each substation and transformer into a centralized control command, which is sent to each edge computing control platform. The reconstruction of each segmented feeder level and transformer level is executed by the corresponding substation edge computing control platform, and the switch action plan is converted into a control command to remotely control each switch. The substation-level reconstruction is executed by the cloud platform, and the switch decision results are converted into centralized control commands to be executed separately by the relevant substation edge nodes.

Images