CN112909942B - Active distribution network fault recovery method and system based on hybrid solution strategy - Google Patents

Abstract

The invention relates to an active distribution network fault recovery method and system based on a hybrid solution strategy. The method includes the following steps: Step S1, constructing an active distribution network island division and network reconstruction that includes distributed power sources and energy storage. model, the objective function of which considers equivalent outage load, segmented switching operation and network loss cost; Step S2, adopt a hybrid solution strategy based on breadth-first search, depth-first search and second-order cone optimization to solve the unified problem The model is solved to obtain a line maintenance plan; step S3 is to realize active distribution network fault recovery based on the line maintenance plan. Compared with the existing technology, the present invention has the advantages of improving the fault recovery rate and solving the maintenance plan quickly.

Description

Active distribution network fault recovery method and system based on hybrid solution strategy

Technical field

The invention relates to an active distribution network fault recovery technology, and in particular to an active distribution network fault recovery method and system based on a hybrid solution strategy.

Background technique

Distribution network failures directly affect social production and residents' daily power supply. In response to sudden natural disasters, instead of protecting the distribution network, the power grid began to look for ways to quickly restore the distribution system using black-start distributed power supplies after the disaster to improve reliability. The active distribution network (AND) has distributed generation (DG) that can be used as a black-start power supply as a backup. Its network topology is flexible and changeable and provides more opportunities for the active distribution network fault recovery strategy. Optimization possible. Therefore, fault recovery strategies for distribution network topology transformation have become one of the research hotspots in the field of distribution network.

After a fault occurs in the active distribution network, the power supply is generally restored by reasonably and effectively dividing the distributed power sources to form an island operation, or by changing the network topology through network reconstruction to connect the remaining power loss areas to the main network and restore power supply. . Compared with traditional distribution networks, fault recovery of active distribution networks containing DG requires more factors to be considered. The intermittency and fluctuation of load and DG output will bring more influencing factors to the original fault recovery method. Existing fault recovery methods also have the problem of insufficient fault recovery rate.

Contents of the invention

The purpose of the present invention is to overcome the above-mentioned shortcomings of the prior art and provide an active distribution network fault recovery method and system based on a hybrid solution strategy that improves the fault recovery rate and quickly solves maintenance plans.

The object of the present invention can be achieved through the following technical solutions:

An active distribution network fault recovery method based on hybrid solution strategy, including the following steps:

Step S1, construct a unified model of active distribution network island division and network reconstruction including distributed power supply and energy storage. The objective function of this model considers equivalent power loss load, segmented switching operation and network loss cost;

Step S2: Use a hybrid solution strategy based on a combination of breadth-first search, depth-first search and second-order cone optimization to solve the unified model and obtain a line maintenance plan;

Step S3: Implement active distribution network fault recovery based on the line maintenance plan.

Further, the objective function is expressed as:

In the formula: g _{1, t} is the network loss of the system in period t; g _{2, t} is the number of segmented switching operations of the system in period t; g _{3, t} is the equivalent load recovery amount of the system in period t; λ _i is node i. The load weight; P _i,t is the active power of load node i in load period t; Represents the set of load nodes; υ ₁ , υ ₂ , and υ ₃ are the network loss cost coefficient, switching operation cost coefficient and equivalent power loss load cost coefficient respectively, N _T is the number of time periods that the entire fault recovery lasts, Δt is a single time The duration of the segment.

Further, the calculation method of the network loss is:

In the formula: I _ij,t is the effective value of the current of branch ij during period t of the system, R _ij is the resistance value of branch ij; Ω is the set of all branches of the active distribution network;

The calculation method for the number of segmented switch operations is:

In the formula: the decision variable α _ij,t is the switch status of line ij in period t. Taking 0 means that the switch of line ij is open, and taking 1 means that the switch of line ij is closed;

The calculation method of the equivalent load recovery amount is:

In the formula: the decision variable y _i,t is the load recovery status of node i in period t. Taking 1 means that load i has recovered in period t, and taking 0 means that load i has not recovered in period t.

Further, the constraints of the unified model of active distribution network islanding and network reconstruction include node voltage and branch current constraints, node power balance constraints, DG power constraints, network structure constraints, energy storage charge and discharge status and power constraints , energy storage remaining capacity constraints, capacitor switching constraints and fault maintenance strategy constraints.

Further, in step S2, solving the unified model is specifically as follows:

Step S201: Read the fault planned power outage time T, DG output power during this period, load forecast power and energy storage state of charge;

Step S202: Traverse each DG and energy storage access node as the root node, use the breadth-first search algorithm to determine the power circle, and obtain a feasible solution to the island;

Step S203: According to the feasible solution of the island, use the depth-first search algorithm to obtain the initial island division range, and then modify the initial island division range according to the island fusion strategy to obtain the island division result for this period;

Step S204: Based on the islanding division results of this period, the position of the operable contact switch and the network topology, obtain the unified model of islanding division and network reconstruction of the active distribution network in this period;

Step S205: Construct intermediate variables, linearize the unified model, and convert it into a standard mixed integer second-order cone model;

Step S206: Use a commercial solver to solve the standard mixed integer second-order cone model;

Step S207: Determine whether the calculations for all time periods have been completed. If so, output the final line maintenance plan. If not, return to step S201 to perform calculations for the next time period.

Further, the modification of the initial island division range according to the island fusion strategy is specifically as follows:

Determine whether there is an intersection among the islands under the initial island division range. If so, merge the islands with intersections, record the equivalent DG and its directly connected equivalent nodes, and return to step S202. If not, determine the adjacent islands. Whether the island satisfies the fusion constraint conditions, if so, merge, otherwise cancel the fusion, and execute step S204.

Further, the commercial solver includes the YALMIP toolbox.

In step S205, a second-order cone relaxation technique is used to perform convex relaxation on the unified model, and the unified model is converted into the standard mixed integer second-order cone model.

Further, the feasible region of the standard mixed integer second-order cone model is relaxed to the entire second-order cone, and the search space is within the range of the convex cone.

The invention also provides an active distribution network fault recovery system based on a hybrid solution strategy, including:

One or more processors, a memory, and one or more programs stored in the memory, the one or more programs including instructions for executing the active distribution network fault recovery method based on a hybrid solution strategy as described above.

Compared with the prior art, the present invention has the following beneficial effects:

1. The present invention constructs a unified model of active distribution network fault recovery that considers island division and network reconstruction, and proposes a corresponding hybrid solution strategy, which can comprehensively consider various influencing factors, jointly complete fault recovery in non-fault outage areas, and improve Reliability of the maintenance plan.

2. The objective function of the present invention considers the equivalent power loss load, segmented switching operation and network loss cost, and can comprehensively consider and integrate the amount of loss during the fault, the economy corresponding to the segmented switching operation and the energy efficiency corresponding to the network loss.

3. The present invention uses a hybrid solution strategy based on a combination of breadth-first search, depth-first search and second-order cone optimization to achieve accurate and efficient solution of the active distribution network fault planning model and improve model solution efficiency.

4. The present invention incorporates capacitor switching constraints, so that the fault recovery model can take into account both active power and reactive power, and improve the fault recovery rate.

5. The present invention incorporates the constraints of the fault maintenance strategy and optimizes the fault repair sequence by adjusting the switching decision variables of the fault line, thereby determining the optimal fault maintenance sequence and improving the power supply recovery rate.

Description of the drawings

Figure 1 shows the steps of the active distribution network fault recovery method according to the present invention;

Figure 2 is the solution flow chart of the unified model of island division and network reconstruction;

Figure 3 shows an IEEE 33-node active distribution network containing DG and energy storage;

Figure 4 shows the fault recovery results of island division and network reconstruction (14:00-15:00);

Figure 5 shows the fault recovery results of island division and network reconstruction (15:00-16:00);

Figure 6 shows the fault recovery results of island division and network reconstruction (16:00-17:00);

Figure 7 shows the fault recovery results of island division and network reconstruction (17:00-18:00);

Figure 8 shows the output results of energy storage ES1 during the fault;

Figure 9 shows the output results of energy storage ES2 during a fault.

Detailed ways

The present invention will be described in detail below with reference to the accompanying drawings and specific embodiments. This embodiment is implemented based on the technical solution of the present invention and provides detailed implementation modes and specific operating procedures. However, the protection scope of the present invention is not limited to the following embodiments.

As shown in Figure 1, the present invention provides an active distribution network fault recovery method based on a hybrid solution strategy, which includes the following steps:

Step S1, construct a unified model of active distribution network island division and network reconstruction including distributed power supply and energy storage. The objective function of this model considers equivalent power loss load, segmented switching operation and network loss cost;

Step S2: Use a hybrid solution strategy based on a combination of breadth-first search, depth-first search and second-order cone optimization to solve the unified model and obtain a line maintenance plan;

Step S3: Implement active distribution network fault recovery based on the line maintenance plan.

In step S1, a fault recovery strategy combining island division and network reconstruction is adopted. After an active distribution network failure occurs, based on the importance of the load, the power supply needs of some important loads are prioritized through island division, and at the same time, the power supply needs of some important loads must be met as much as possible. Avoid the generation of a large number of unrecoverable nodes after island division, that is, nodes that cannot restore power supply through network reconstruction; then, based on the network topology after island division, use network reconstruction to restore power supply to remaining load nodes, optimizing system power flow and network losses , while also minimizing the number of switch operations. The objective function of the established unified model of active distribution network islanding and network reconstruction is to minimize the comprehensive operating cost during active distribution network faults, which takes into account the equivalent power loss load, segmented switching operations and network loss costs. The objective function is specific. Expressed as:

In the formula: g _{1, t} is the network loss of the system in period t; g _{2, t} is the number of segmented switching operations of the system in period t; g _{3, t} is the equivalent load recovery amount of the system in period t; λ _i is node i. The load weight; P _i,t is the active power of load node i in load period t; Represents the set of load nodes; υ ₁ , υ ₂ , and υ ₃ are the network loss cost coefficient, switching operation cost coefficient and equivalent power loss load cost coefficient respectively, N _T is the number of time periods that the entire fault recovery lasts, Δt is a single time The duration of the segment.

The calculation method of the network loss is:

In the formula: I _ij,t is the effective value of the current of branch ij during period t of the system, R _ij is the resistance value of branch ij; Ω is the set of all branches of the active distribution network.

The calculation method for the number of segmented switch operations is:

In the formula: the decision variable α _ij,t is the switch status of line ij in period t. Taking 0 means that the switch of line ij is open, and taking 1 means that the switch of line ij is closed;

The calculation method of the equivalent load recovery amount is:

In the formula: the decision variable y _i,t is the load recovery status of node i in period t. Taking 1 means that load i has recovered in period t, and taking 0 means that load i has not recovered in period t.

The constraints of the unified model of active distribution network islanding and network reconstruction include node voltage and branch current constraints, node power balance constraints, DG power constraints, network structure constraints, energy storage charge and discharge status and power constraints, and energy storage remaining capacity. constraints, capacitor switching constraints and troubleshooting strategy constraints.

(1) Node voltage and branch current constraints

When the load is restored to the grid through islanding or network reconstruction, to ensure that the system can operate normally and stably, the node voltage and branch current in the network need to meet certain constraints, namely:

In the formula: V _i ^max and V _i ^min represent the upper and lower limits of the voltage of node i respectively; I _ij,t is the current flowing on branch ij during t period; is the maximum current allowed to flow through branch ij.

(2) Node power balance constraints

Meeting the requirements of power balance is the key to the stable operation of the active distribution network after islanding or network reconstruction. It is easy to know from Kirchhoff's law that the sum of the power flowing into a node must be equal to the sum of the power flowing out of the node, so the power balance constraint of the node should be satisfied, that is:

In the formula: V _i,t and V _j,t are the voltages of nodes i and j in period t; G _ij and B _ij are the conductance and susceptance of branch ij respectively; δ _ij,t is the voltage phase angle of branch ij in period t Difference; C(i) is the set of nodes connected to node i.

(3)DG power constraints

Since the output of wind turbine (WT) and photovoltaic (PV) is highly intermittent and fluctuating, in order for the system to operate stably after island division and network reconstruction, the DG power constraints should be met, namely:

In the formula: P _DG,i,t and Q _DG,i,t are the active power output and reactive power output of DG at node i in period t; is the upper limit of active output and reactive output of DG at node i during period t;/> It is the lower limit of DG active output and reactive output at node i in period t.

(4)Network structure constraints

A fault recovery strategy that considers island division and network reconstruction. During the fault recovery process, the active distribution network needs to satisfy connectivity constraints and radial constraints, namely:

In the formula: f _di is the virtual load of node i. The function of virtual load is to ensure that all load nodes meet the connectivity constraints during the fault recovery process. If a load node is isolated, the virtual power balance constraint of the node is not satisfied, that is Therefore, if the node virtual power balance constraint needs to be satisfied, all load nodes need to remain connected. At the same time, since the number of network line connections in the distribution network is equal to N-1, N is the number of nodes, satisfying the connectivity constraint means satisfying the radial constraint, so the virtual The role of load and virtual traffic is to ensure that network connectivity constraints and radial constraints are met. Generally, the unit is 1, f _ij,t is the virtual traffic flowing through branch ij during t period, N _b is the number of branches, N _n is the number of nodes, and N _s is the number of power supplies.

(5) Energy storage charge and discharge status and power constraints

Energy storage devices can be used in the process of island division and network reconstruction, but the charge and discharge power of the energy storage should not be greater than its limit value, so the energy storage charge and discharge state and power constraints need to be met, namely:

In the formula: the decision variables are the energy storage charging and discharging state y _e,i,t and the energy storage charging and discharging power P _e,i,t respectively. y _e,i,t represents the 0-1 variable of the charging and discharging state of the energy storage at node i during period t, with 0 representing charging and 1 representing discharging; Represents the maximum charging and discharging power of energy storage at node i respectively; P _{e, i, t} represents the charging or discharging power of energy storage at node i in period t.

(6) Energy storage remaining capacity constraints

Energy storage has certain capacity limits and cannot be overcharged or discharged. It should satisfy the remaining capacity constraints of energy storage, namely:

In the formula: is the remaining capacity of energy storage at node i during period t;/> and/> The maximum and minimum capacity limits of energy storage at node i; eta _ch and eta _dis are the charging and discharging efficiency of energy storage respectively.

(7) Capacitor switching constraints

In order to cope with the reactive power demand of the load during islanding operation of the active distribution network and the undervoltage problem caused by reactive power shortage after network reconstruction, capacitor banks need to be switched for reactive power compensation during the islanding division and network reconstruction fault recovery process. The capacitor switching constraints should be met, that is:

In the formula: is the reactive power compensation capacity of the capacitor at node i during period t;/> Indicates the reactive power compensation capacity invested in a single capacitor; decision variable/> The number of capacitors invested in node i during period t;/> is the total number of capacitors available for input at node i.

(8) Troubleshooting strategy constraints

When multiple lines in the active distribution network fail, it is necessary to reasonably arrange the maintenance sequence of each line fault within the planned power outage period. The specific method is to switch each fault line on and off in a unified model of network reconstruction and island division. The state variable imposes some constraints, that is: the total number of closed switches in the line in the next period is always more than the total number of switches in the previous period, but it cannot exceed the maximum number of lines that can be repaired in each period. When the islanding division and network reuse in each period are After the construction is completed, by optimizing the decision variables of each faulty line switch, the remaining faulty lines are repaired, the topology of the network is indirectly adjusted and optimized, and the optimal fault maintenance strategy for the line during the entire fault period is finally determined.

In the formula: the decision variable β _ij,t represents the status of each fault line switch; Ω _E is the set of all fault lines; k is the maximum number of lines that can be repaired in a single period.

As shown in Figure 2, in step S2, the specific steps to solve the unified model are:

Step S201: Read the fault planned power outage time T, DG output power during this period, load forecast power and energy storage state of charge;

Step S202: Traverse each DG and energy storage access node as the root node, use the breadth-first search algorithm to determine the power circle, and obtain a feasible solution to the island;

Step S203: According to the feasible solution of the island, use the depth-first search algorithm to obtain the initial island division range, and then modify the initial island division range according to the island fusion strategy to obtain the island division result for this period;

The modification of the initial island division range according to the island fusion strategy is specifically: determining whether there is an intersection among the islands under the initial island division range, and if so, merging the islands with intersections and recording equivalent DGs and their directly connected For equivalent nodes, return to step S202. If not, determine whether the adjacent islands meet the fusion constraints. If so, merge; otherwise, cancel the fusion and execute step S204;

Step S204: Based on the islanding division results of this period, the position of the operable contact switch and the network topology, obtain the unified model of islanding division and network reconstruction of the active distribution network in this period;

Step S205: Construct intermediate variables, linearize the unified model, and convert it into a standard mixed integer second-order cone model;

Step S206: Use a commercial solver to solve the standard mixed integer second-order cone model. Commercial solvers that can be used include the MOSEK algorithm package in the YALMIP toolbox, etc.;

Step S207: Determine whether the calculations for all time periods have been completed. If so, output the final line maintenance plan. If not, return to step S201 to perform calculations for the next time period.

In step S205, there are a large number of quadratic terms and trigonometric function terms in the unified model of island partitioning and network reconstruction, which is a MINLP problem. The second-order cone relaxation technology is used to perform convex relaxation on the model. The quadratic terms and trigonometric function terms existing in the unified model of active distribution network islanding and network reconstruction, as shown in equations (4) and (6), are non-convex and difficult to solve directly mathematically. By introducing intermediate variables, the present invention uses second-order cone relaxation technology to perform convex relaxation processing on the unified model of island division and network reconstruction, and further can use mature commercial solvers to quickly obtain the global optimal solution. The introduced intermediate variables are shown in formulas (13) to (17).

C _ij,t =V _i,t V _j,t cos(δ _ij,t ) (15)

D _ij,t =V _i,t V _j,t sin(δ _ij,t ) (16)

Factor (17) contains quadratic terms, and the model still has non-convexity. After further relaxing it, it is shown in equation (18).

After deformation of equation (18), its second-order cone form can be obtained as shown in equation (19).

The network loss equation (4) containing square terms and the node power balance constraint equation (6) containing square terms and trigonometric function terms in the original model, after second-order cone relaxation, are shown in equations (20) to (22) respectively. .

At this time, the feasible region is relaxed to the entire second-order cone, and the search space is limited to the convex cone range. It has a convex feasible region and can be solved using a commercial solver.

If the above functions are implemented in the form of software functional units and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention essentially or the part that contributes to the existing technology or the part of the technical solution can be embodied in the form of a software product. The computer software product is stored in a storage medium, including Several instructions are used to cause a computer device (which may be a personal computer, a server, or a network device, etc.) to execute all or part of the steps of the methods described in various embodiments of the present invention. The aforementioned storage media include: U disk, mobile hard disk, read-only memory (ROM, Read-Only Memory), random access memory (RAM, Random Access Memory), magnetic disk or optical disk and other media that can store program code. .

In another embodiment, an active distribution network fault recovery system based on a hybrid solution strategy is provided, including: one or more processors, a memory, and one or more programs stored in the memory, the one or more Each program includes instructions for executing the active distribution grid fault recovery method based on a hybrid solution strategy as described above.

Example

This embodiment uses an IEEE 33-node active distribution network containing DG and energy storage (ES). Its structure is shown in Figure 3, and the node numbers are marked in the figure; the maximum load of the system is 3715kW+2300kvar. The load level parameters are shown in Table 1. The specific access nodes and rated power of each DG and energy storage are shown in Table 2 and Table 3.

Table 1 Node load level parameters

Table 2 DG rated power and access points

Table 3 Energy storage access points and parameters

In order to verify the effectiveness of the above active distribution network fault recovery method, it is assumed that under extreme weather conditions, when a major accident occurs in the distribution network where multiple lines fail simultaneously, lines 6-7, 10-11, 15-16, and 22 are set up. A permanent fault occurs on -23 and 26-27, and the power outage period is 14:00-18:00. This embodiment stipulates that the maximum number of maintenance lines in a single period is 1. After the island division and network reconstruction in each period are completed, on the premise of satisfying the fault maintenance strategy constraints of equation (12), the fault repair is performed by adjusting the switching decision variables of the fault line. Optimize in sequence to determine the optimal troubleshooting strategy. The main goal of the fault recovery strategy that combines island division and network reconstruction proposed by the present invention is to restore power supply to as many loads as possible, and the remaining objective functions are used to assist in optimizing the network operating status.

Adopting the fault recovery strategy proposed by the present invention that combines island division and network reconstruction, and considering the constraints of the fault maintenance strategy, before performing island division and network reconstruction operations in each period, the switching decision variables of the faulty line are first optimized and adjusted. After adjusting the network topology, the final line maintenance strategies are 22-23, 10-11, 15-16, and 6-7. The fault recovery results in each period are shown in Figures 4 to 9 and Table 4. It can be seen that ES1 has a larger rated power and cooperates with WT2 to participate in island division, while ES2 can only play a role in network reconstruction due to its small rated capacity. Class 1 load nodes 8 and 8 can be realized through the method of the present invention. Power supply will be restored on 9, 16, 24, 30 and 31.

Table 4 Fault recovery results in each period

In order to verify the superiority of the fault recovery strategy proposed in the present invention, the operation results are compared with the operation results of the islanding division and network reconstruction without cooperation but taking into account the fault maintenance strategy, and the islanding division and network reconstruction cooperation but no fault maintenance strategy. The results are shown in Table 5.

Table 5 Comparison of fault recovery strategies

It can be seen that using strategy 1, that is, island division and network reconstruction do not cooperate with each other. Although the load restored through island division reaches the maximum, because the island division and network reconstruction are performed independently, only the island division is obtained. The local optimal solution resulted in the failure of some nodes to restore power supply even through network reconstruction, and the power supply recovery rate was only 87.15%. Strategy 2 is adopted, that is, island division and network reconstruction are combined to restore power supply, but there is no fault maintenance strategy. The order of repairing lines is random, and the power supply recovery rate is 92.43%. Comparing strategy 2 and strategy 1, it can be concluded that the fault recovery of the active distribution network should be carried out by considering the combination of island division and network reconstruction. Island division and network reconstruction are carried out at the same time, even if the reconstruction process is considered, the failure cannot be restored. load, the result shows that the power loss load has decreased by 508.7kW·h, which proves the superiority of the unified model of island division and network reconstruction in the present invention. When strategy 3 is adopted, that is, island division and network reconstruction are combined to restore power supply and fault maintenance strategies are taken into consideration, the power supply recovery rate can reach 97.27%, and the power loss load is reduced by 974.4kW·h and 465.7kW· respectively compared with strategies 1 and 2. h, this shows that through the fault maintenance strategy adopted in the present invention, the fault repair process can be arranged more reasonably, thereby further improving the power supply recovery rate.

The preferred embodiments of the present invention are described in detail above. It should be understood that those skilled in the art can make many modifications and changes based on the concept of the present invention without creative efforts. Therefore, any technical solutions that can be obtained by those skilled in the art through logical analysis, reasoning or limited experiments based on the concept of the present invention and on the basis of the prior art should be within the scope of protection determined by the claims.

Claims (8)

1. An active distribution network fault recovery method based on a hybrid solution strategy, which is characterized by including the following steps: Step S1, construct a unified model of active distribution network island division and network reconstruction including distributed power supply and energy storage. The objective function of this model considers equivalent power loss load, segmented switching operation and network loss cost; Step S2: Use a hybrid solution strategy based on a combination of breadth-first search, depth-first search and second-order cone optimization to solve the unified model and obtain a line maintenance plan; Step S3: Implement active distribution network fault recovery based on the line maintenance plan; The objective function is expressed as: In the formula: g _{1, t} is the network loss of the system in period t; g _{2, t} is the number of segmented switching operations of the system in period t; g _{3, t} is the equivalent load recovery amount of the system in period t; λ _i is node i. The load weight; P _i,t is the active power of load node i in load period t; Represents the set of load nodes; υ ₁ , υ ₂ , and υ ₃ are the network loss cost coefficient, switching operation cost coefficient and equivalent power loss load cost coefficient respectively, N _T is the number of time periods that the entire fault recovery lasts, Δt is a single time The duration of the segment; The calculation method of the network loss is: In the formula: I _ij,t is the effective value of the current of branch ij during period t of the system, R _ij is the resistance value of branch ij; Ω is the set of all branches of the active distribution network; The calculation method for the number of segmented switch operations is: In the formula: the decision variable α _ij,t is the switch status of line ij in period t. Taking 0 means that the switch of line ij is open, and taking 1 means that the switch of line ij is closed; The calculation method of the equivalent load recovery amount is: In the formula: the decision variable y _i,t is the load recovery status of node i in period t. Taking 1 means that load i has recovered in period t, and taking 0 means that load i has not recovered in period t. 2. The active distribution network fault recovery method based on a hybrid solution strategy according to claim 1, characterized in that the constraints of the unified model of islanding division and network reconstruction of the active distribution network include node voltages and branch currents. Constraints, node power balance constraints, DG power constraints, network structure constraints, energy storage charge and discharge status and power constraints, energy storage remaining capacity constraints, capacitor switching constraints and fault maintenance strategy constraints. 3. The active distribution network fault recovery method based on hybrid solution strategy according to claim 1, characterized in that, in step S2, solving the unified model is specifically: Step S201: Read the fault planned power outage time T, DG output power during this period, load forecast power and energy storage state of charge; Step S202: Traverse each DG and energy storage access node as the root node, use the breadth-first search algorithm to determine the power circle, and obtain a feasible solution to the island; Step S203: According to the feasible solution of the island, use the depth-first search algorithm to obtain the initial island division range, and then modify the initial island division range according to the island fusion strategy to obtain the island division result for this period; Step S204: Based on the islanding division results of this period, the position of the operable contact switch and the network topology, obtain the unified model of islanding division and network reconstruction of the active distribution network in this period; Step S205: Construct intermediate variables, linearize the unified model, and convert it into a standard mixed integer second-order cone model; Step S206: Use a commercial solver to solve the standard mixed integer second-order cone model; Step S207: Determine whether the calculations for all time periods have been completed. If so, output the final line maintenance plan. If not, return to step S201 to perform calculations for the next time period. 4. The active distribution network fault recovery method based on hybrid solution strategy according to claim 3, characterized in that the modification of the initial island division range according to the island fusion strategy is specifically: Determine whether there is an intersection among the islands under the initial island division range. If so, merge the islands with intersections, record the equivalent DG and its directly connected equivalent nodes, and return to step S202. If not, determine the adjacent islands. Whether the island satisfies the fusion constraint conditions, if so, merge, otherwise cancel the fusion, and execute step S204. 5. The active distribution network fault recovery method based on hybrid solving strategy according to claim 3, characterized in that the commercial solver includes YALMIP toolbox. 6. The active distribution network fault recovery method based on hybrid solution strategy according to claim 3, characterized in that, in step S205, the second-order cone relaxation technology is used to perform convex relaxation on the unified model, and the unified model is converted into The standard mixed integer second-order cone model. 7. The active distribution network fault recovery method based on a hybrid solution strategy according to claim 3, characterized in that the feasible region of the standard mixed integer second-order cone model is relaxed to the entire second-order cone, and the search space is Inside the convex cone range. 8. An active distribution network fault recovery system based on a hybrid solution strategy, which is characterized by including: One or more processors, a memory, and one or more programs stored in the memory, the one or more programs including a program for performing active power distribution based on a hybrid solution strategy as described in any one of claims 1-7 Instructions for network failure recovery methods.