Petal type power distribution network fault self-healing method based on double-layer optimization model

Abstract

The invention relates to a petal-type power distribution network fault self-healing method based on a double-layer optimization model. The method comprises the following steps: S1, analyzing a PIES access petal network fault blocking mechanism; s2, establishing a double-layer optimization model based on'reconstruction-excitation '; s3, solving a lower layer optimization model based on the excitation type DR; and S4, designing a petal net dynamic fault self-healing strategy process based on the double-layer optimization model. According to the method, rich contact resources of the petal net are fully utilized, and the economical efficiency of the motivation type DR is improved; the potential that PIES participates in fault self-healing by utilizing the advantages of multi-energy complementary power supply is fully excavated, the benefits of the petal net and the PIES are balanced through a master-slave game, and the coupling relationship among multiple subjects is enhanced; the double-layer optimization model complements each other, is suitable for various fault scenes of the petal network, effectively eliminates fault blocking, gives consideration to reliability and economical efficiency, and has certain practical engineering value due to the feasibility, the optimal configuration of PIES energy in the space-time dimension and the uninterrupted continuous operation of important loads.

Description

Technical Field

[0001] The present invention belongs to the technical field of distribution network fault self-healing, and in particular relates to a petal-type distribution network fault self-healing method based on a double-layer optimization model. Background Art

[0002] In recent years, the access of park-level integrated energy system (PIES) to distribution network has been widely studied and applied. PIES is a terminal energy supply system directly facing users, which conducts two-way power interaction with the distribution network and can be regarded as an emergency power supply to support the self-healing of distribution network faults. In order to cope with the high proportion of PIES access and improve the reliability of power supply, Xiongan New Area introduced a multi-segment and moderately connected design and built a highly reliable and flexible double-petal distribution network (hereinafter referred to as the petal network). With the vigorous construction and promotion of the petal network, it will become an important carrier for PIES access. Therefore, it is of great significance to study the fault self-healing strategy of PIES access to the petal network.

[0003] Traditional distribution network fault self-healing strategies only take independent power grids as research objects, and use network reconstruction and island division to restore and control faulty distribution networks. In addition, existing strategies are all based on radial distribution networks with "closed-loop design and open-loop operation". Their objective functions and constraints are difficult to directly apply to petal networks. At present, there is a lack of research on fault self-healing strategies for new grid-structured distribution networks such as petal networks. On the one hand, the special closed-loop operation mode and the access of a considerable number of PIES make the power flow of the petal network complex, multi-source and bidirectional, which is quite different from the radial network; on the other hand, the topological structure of the petal network involves two substation nodes and multiple switch station nodes, with rich inter-station connections, and many potential topological combinations and transfer paths. However, after the fault isolation of the petal network, the normal line may be overloaded or overloaded, resulting in fault blocking. Fault blocking will not only increase network loss and shorten cable life, but also may cause line chain failures, affecting power supply reliability. For this reason, petal network fault self-healing requires the removal of some loads and the coordination of PIES. According to relevant research, petal networks, PIES, and interruptible loads (IL) belong to different interest entities. However, the existing strategies all have fixed compensation prices for PIES or ILs that participate in fault self-healing, and a complete incentive or compensation mechanism has not yet been established, resulting in conflicts of interest among multiple subjects, affecting the reliability and economy of fault self-healing. Therefore, it is urgent to study a new fault self-healing strategy to balance the distribution of interests between the petal network and PIES, and strengthen the coupling relationship between multiple subjects.

[0004] In order to eliminate the fault blocking of the Petal Network and realize the coordinated optimization operation of the Petal Network and PIES, the present invention proposes a fault self-healing strategy based on a double-layer optimization model. Summary of the invention

[0005] The purpose of the present invention is to propose a petal-type distribution network fault self-healing method based on a double-layer optimization model in view of the deficiencies in the prior art. With the help of big data technology and artificial intelligence technology, the petal network and PIES can deeply perceive and integrate and interact, making it more feasible for PIES to quickly support the petal network fault self-healing and optimized operation; the method uses the distribution management system (distribution management system, DMS) of the petal network and the distribution system operator (distribution system operator, DSO) to hierarchically solve the upper and lower optimization models respectively, so as to improve the efficiency and speed of solving the fault self-healing model; in addition, the method uses incentive-type demand response (demand response, DR) to perform emergency control of PIES and IL, introduces the master-slave game idea, and builds an information and decision-making interface between the petal network and PIES through the incentive electricity price, so as to realize the coordinated optimization and benefit balance between the petal network and PIES, and effectively make up for the deficiencies of the existing methods.

[0006] The present invention solves the technical problem by the following technical solutions:

[0007] A petal-type distribution network fault self-healing method based on a double-layer optimization model, the steps of the method are:

[0008] S1. Analyze the mechanism of PIES access to the petal network fault blocking, weight the power transfer entropy by the inverse of the line load rate, and obtain the fault-weighted power transfer entropy of the petal network. Use the "entropy change" process of the power flow after the fault to quantify the probability of the petal network line blocking and the severity of the fault blocking as the theoretical basis for solving the double-layer optimization model.

[0009] S2. Considering the reliability and economy of the self-healing of the petal network fault, a two-layer optimization model based on "reconstruction-incentive" is established. The DMS uses the switch state of the petal network as the decision variable to solve the upper-layer optimization model based on network reconstruction, including the objective functions of load balancing, network loss and switch operation simplicity, as well as the constraints of network topology constraints, line flow constraints, and voltage and current constraints, as the basis for solving the lower-layer optimization model;

[0010] S3, DSO uses the load shedding power of the petal network and the additional power of PIES as decision variables to solve the lower-level optimization model based on incentive-based DR. The main player is DSO, the game strategy is the incentive electricity price and the load shedding power of the petal network, the game benefit is the comprehensive objective function of the load balancing degree of the petal network and the fault dispatching cost, and the game constraints are line flow constraints, voltage and current constraints, incentive electricity price constraints and load shedding constraints; the slave player is the PIES aggregator, the game strategy is the PIES energy equipment output and energy trading plan, the game benefit is the PIES operating income, and the game constraints are energy balance constraints, energy equipment output constraints and energy trading constraints; finally, DSO and PIES aggregator realize game interaction and collaborative optimization through incentive electricity prices, among which DSO considers multi-dimensional factors to compensate for the load shedding losses of IL users, and PIES aggregator optimizes and adjusts the interactive power between PIES and petal network;

[0011] S4. Design a dynamic fault self-healing strategy process for the petal network based on a two-layer optimization model, divide the petal network fault repair phase into multiple discrete time periods, and dynamically adjust the fault self-healing decision.

[0012] Furthermore, the S4 is specifically as follows: when a failure occurs in the petal network, the steps of the strategy are as follows:

[0013] 1) DMS collects the operating parameters and switch status of the petal network in real time, quickly locates and isolates the faulty components based on the fault information and operating data, and completes the load transfer;

[0014] 2) DMS determines whether there is a blocked line in the petal network based on the line load rate information. If a fault blockage occurs, proceed to the next step; otherwise, go to 7);

[0015] 3) DMS solves the upper optimization model, performs topology search based on the BFS algorithm, and quickly calculates and generates the optimal network reconstruction plan under the current load power and PIES interaction power by taking into account the objective function and constraints established by S2, sends the switch state conversion instruction to the switch station, and simultaneously uploads the network reconstruction plan to the DSO as the basis for solving the lower optimization model;

[0016] 4) If the upper layer optimization model has no solution, that is, the network reconstruction fails, go directly to 6); otherwise, go to the next step;

[0017] 5) DMS determines whether the fault blocking is eliminated. If there is still a blocked line, it goes to the next step; otherwise, it goes to 7);

[0018] 6) DSO solves the lower-level optimization model, conducts a master-slave game with PIES aggregators through incentive electricity prices, considers multi-dimensional factors to compensate for the load shedding losses of IL users, and sends the optimal load shedding plan to DMS, which executes the load shedding instructions; PIES aggregators formulate the optimal operation plan of PIES, perform energy management on related PIES, and optimize and adjust the interaction power between PIES and Petal Network;

[0019] 7) DMS determines whether a new fault occurs in the Petal Network. If so, it restarts the fault self-healing strategy; otherwise, DMS determines that the Petal Network is operating safely and stably during the current period.

[0020] In order to eliminate the blockage of petal network faults and realize rapid self-healing and optimized operation after petal network faults, the present invention proposes a fault self-healing strategy based on a double-layer optimization model. Compared with the prior art, the present invention can produce the following positive effects:

[0021] Firstly, the upper-layer optimization model based on network reconstruction makes full use of the rich communication resources of the Petal Network to improve the economy of incentive-based DR; secondly, the lower-layer optimization model based on incentive-based DR fully explores the potential of PIES to participate in fault self-healing by taking advantage of the advantages of multi-energy complementary power supply, balances the interests of the Petal Network and PIES through master-slave game, and enhances the coupling relationship between multiple subjects; finally, the two-layer optimization model complements each other and is suitable for various fault scenarios of the Petal Network, effectively eliminating fault blocking and taking into account both reliability and economy. The dynamic fault self-healing strategy of the Petal Network based on the two-layer optimization model ensures the feasibility of fault self-healing decision-making, the optimal configuration of PIES energy in the time and space dimensions, and the continuous operation of important loads without power outages, which has certain practical engineering value. BRIEF DESCRIPTION OF THE DRAWINGS

[0022] Figure 1 This is the operation structure diagram of the PIES of the present invention connected to the 10kV petal network;

[0023] Figure 2 This is a framework diagram of a two-layer optimization model based on "reconstruction-incentive" of the present invention;

[0024] Figure 3 This is a topology search flow chart based on the BFS algorithm of the present invention;

[0025] Figure 4 It is a schematic diagram of the PIES structure of the present invention and its interaction power with the petal network;

[0026] Figure 5 This is a framework diagram of the lower layer optimization model based on incentive-based DR of the present invention;

[0027] Figure 6 It is the action time diagram of each stage of the dynamic fault self-healing strategy of the petal network of the present invention;

[0028] Figure 7 This is a flow chart of the petal network fault self-healing strategy based on the double-layer optimization model of the present invention. DETAILED DESCRIPTION

[0029] The present invention is further described in detail below through specific examples. The following examples are only illustrative and not restrictive, and the protection scope of the present invention cannot be limited thereto.

[0030] The present invention provides a petal-type distribution network fault self-healing method based on a double-layer optimization model, comprising:

[0031] 1. Analysis of the mechanism of PIES access to Huaban Network failure and blocking

[0032] Figure 1 This is the operating structure of PIES access to the 10kV petal network. The characteristics of petal network fault blocking are analyzed below, and the probability of fault blocking in the petal network and the severity of fault blocking are quantified by fault weighted power flow transfer entropy.

[0033] With the high proportion of PIES access, load demand and cable laying length, normal lines may be blocked after a petal network fault. The reason is that fault isolation changes the topology of the petal network, resulting in an unbalanced distribution of fault transfer power, that is, the load power in the current period is greater than the actual transmission capacity of the line. Since the power supply depth of the petal network is shallow, the possibility of voltage exceeding the limit is very low, and fault blocking is usually manifested as line overload or the presence of overloaded lines close to the power limit.

[0034] The present invention evaluates the reliability and safety of the petal network line through the line load rate, and defines the load rate of the petal network line m at time t as for:

[0035]

[0036] Where: are the actual transmission power and rated transmission power of line m respectively.

[0037] when When , DMS determines that line m is overloaded; when When , DMS determines that line m is overloaded.

[0038] The present invention measures the balance degree of the temporal and spatial distribution of the load of the petal network through the load balance degree. Define the load balance degree of the petal network at time t for:

[0039]

[0040] Where: L is the set of normal lines of the petal network; They are respectively the highest and lowest line load rates in the petal network.

[0041] The smaller it is, the more balanced the temporal and spatial distribution of the petal network load or the line flow distribution is.

[0042] Considering that the petal network is an energy balance system, the present invention quantifies the probability of line blockage through the "entropy change" process of the power flow after the fault. Define the power flow increment of the normal line m after the line n fails at time t as The current impact rate for:

[0043]

[0044] Define the power flow transfer entropy of line m for:

[0045]

[0046] The inverse of the line m load factor is used Weighted, the fault weighted power flow transfer entropy of line n is obtained for:

[0047]

[0048] Where: is the load rate of the normal line k after the line n fails.

[0049] The larger the value, the easier it is for line k to be blocked by a fault. According to the basic entropy theory, the flow increment of all normal lines is evenly distributed according to the load rate, that is, the larger the load rate, the smaller the flow increment of the line. Taking the maximum value, this is the state where the fault transfer flow distribution is most balanced, and at this time, the probability of the petal network line being blocked is the lowest. The smaller the value, the more unbalanced the fault transfer flow distribution is, and the higher the probability of line blocking is. The smaller it is, the higher the load factor of the blocked line will be.

[0050] The topological structure of the petal network after a fault determines the power flow distribution and load power supply path in its non-fault section, and is the main factor causing congestion. Due to the limitation of the topological structure, the efficiency and cost of eliminating congestion by the incentive mechanism are low; and relying solely on network reconstruction may not meet the power flow safety constraints of all lines, resulting in network reconstruction failure and inability to completely eliminate congestion. The petal network cannot operate with blocked lines for a long time, and some loads must be cut off to reduce the load rate of blocked lines. Therefore, in order to improve the reliability and economy of fault self-healing, the present invention establishes a two-layer optimization model based on "reconstruction-incentive", such as Figure 2 shown.

[0051] 2. Upper-layer optimization model based on network reconstruction

[0052] The upper optimization model is that DMS uses the switch status of the Petal Network as a decision variable to reconstruct the network. DMS is the management platform of the Petal Network. The DMS master station calculates and analyzes the fault information and operation data collected and uploaded by the distribution terminal in the Petal Network switch station, and monitors and controls the operation status of the Petal Network in real time.

[0053] 2.1 Objective Function

[0054] 1) Minimum load balancing

[0055]

[0056] Where: is the load balancing degree of Huaban Network.

[0057] 2) Minimum network loss

[0058]

[0059] Where: r ij are the current amplitude and equivalent resistance of line i→j respectively.

[0060] 3) Minimum number of switching actions

[0061]

[0062] Where: F and T are the sets of section switches and tie switches respectively; Indicates the state of the section switch, closed is 1, open is 0; Indicates the state of the tie switch, open is 0 and closed is 1.

[0063] The present invention uses the adaptive weight method to perform dimensionless normalization on the three sub-objective functions and transforms them into a single objective function model F1 as the objective function of network reconstruction:

[0064]

[0065] Where: f i max 、f i min The target f i The maximum and minimum values ​​that can be obtained when solving a single objective function; λ i The target f i The weight can be adaptively adjusted according to the actual fault situation.

[0066] 2.2 Constraints

[0067] 1) Network topology constraints

[0068] The reliability of PIES independent power supply is not as good as the main power supply of the system. In order to improve the power consumption comfort of users and ensure the connectivity between all load nodes and substation nodes, PIES island operation is not considered. If there is a connection path between two substation nodes, long-term operation will form an electromagnetic ring network, causing "shunt phenomenon" and destroying the thermal stability of the system. Therefore, it is necessary to ensure the non-connectivity between substation nodes.

[0069]

[0070] Where: Indicates the opening and closing state of line i→j, the line is 1 when it is on and 0 when it is off; N p is the set of load nodes; n p is the number of load nodes; Respectively represent the connectivity between nodes i, j and substations A, B. If there is a connected path, it is 1, otherwise it is 0; N A 、N B They are respectively the sets of all load nodes in the petal network that have connection paths with substations A and B.

[0071] 2) Line flow constraints

[0072] Based on the Distflow tidal current equation, the present invention uses the second-order cone relaxation method in combination with the "Big-M" method to transform the PIES access to the petal network operation model into a mixed integer second-order cone linear programming model. and Substitute the square of node voltage amplitude and line current amplitude respectively. Introduce auxiliary variables The node voltage drop and the line opening and closing state Get in touch.

[0073]

[0074] Where: are the rated active and reactive power of line i→j respectively; rij 、x ij are the equivalent resistance and reactance of line i→j respectively; are the actual active and reactive power of line i→j respectively; are the active and reactive load powers of node j respectively; are the active and reactive powers injected into node j by PIES respectively; M is a sufficiently large positive real number; ||·||2 represents the 2-norm.

[0075] Equations (11), (12) and (13) are the power flow balance constraints, power flow safety constraints and Ohm's law constraints of the petal network line, respectively. Equation (15) is an equivalent transformation of Equation (14), which obtains the standard second-order cone form of the petal network voltage, current and power, and transforms the model into a convex linear feasible domain, which is convenient for solving with the help of the commercial solver CPLEX and the improved binary particle swarm optimization (PSO) algorithm.

[0076] 3) Voltage and current constraints

[0077]

[0078] Where: v i,max 、v i,min are the squares of the upper and lower limits of the voltage amplitude at node i, respectively; ij,max is the square of the maximum current amplitude of line i→j.

[0079] 2.3 Network topology search method

[0080] In order to improve the efficiency of model solving, the present invention uses the breadth first search algorithm (BFS) to quickly identify and remove solutions that do not meet the network topology constraints in the running model. The BFS algorithm has the idea of ​​"hierarchical search". The topology search process based on the BFS algorithm is as follows: Figure 3 As shown. First, start the search with a substation node as the starting node. If another substation node appears in any of the searched child nodes, it means that there is a connected path between the two substation nodes and the topology is not feasible; otherwise, search again with another substation node as the starting node. After two searches, if there are still untraversed load nodes in the petal network, it means that there are islands and the topology is not feasible; otherwise, the topology is feasible and the judgment result is output.

[0081] 3. Lower-layer optimization model based on incentive-based DR

[0082] PIES has the advantage of multi-energy complementary power supply, and can provide stable and reliable power support for the Petal Network during its fault self-healing period. The structure of PIES and the process of its power interaction with the Petal Network are shown in the figure. Figure 4 As shown. The main energy equipment inside PIES includes photovoltaic (PV), wind turbine (WT), electricity storage (ES), heat storage (HS), gas storage (GS), combined heat and power (CHP) units, power to gas (P2G), electric boiler (EB), etc. The present invention extends the concept of PIES aggregator on the basis of load aggregator, which has basic energy management functions, can uniformly dispatch and manage each PIES connected to the petal network, and optimize and control the interactive power between PIES and the petal network.

[0083] The lower-level optimization model is that DSO uses load shedding power and PIES additional power as decision variables to perform incentive DR on PIES and IL. As the operator of the Petal Network, DSO is responsible for power trading with PIES aggregators. Power trading during Petal Network failures emphasizes speed and efficiency, and DSO needs to use incentive mechanisms to adjust power trading plans. PIES aggregators respond to DSO's incentive DR based on a pre-signed agreement, and increase PIES power sales or reduce PIES power purchases to support fault self-healing without harming their own interests.

[0084] The present invention introduces the master-slave game theory and establishes Figure 5 The lower optimization model shown in the figure. Among them, the main player is DSO, which occupies a dominant position in the game. DSO formulates incentive electricity prices and load shedding plans according to its own needs, and passes the incentive electricity prices to the slave players. The slave players are PIES aggregators, which refer to the incentive electricity prices to formulate PIES operation plans that are in line with their own interests, and feed back the PIES additional power plan to DSO. Based on this plan, DSO makes a new round of adjustments to the incentive electricity price, and conducts multiple rounds of games with PIES aggregators, and finally achieves a constrained balance of interests between the two parties and completes the short-term adjustment of power transactions. In this process, neither party can obtain the other party's internal equipment operation data and user information. This non-invasive trading model can protect the privacy and legitimate interests of the trading subjects.

[0085] 3.1. Agent Game Model

[0086] (1) Game strategies and game benefits

[0087] The main game strategy s1 is the incentive electricity price and petal network load shedding plan formulated by DSO:

[0088]

[0089] Where: is the incentive electricity price of PIES in region q at time t; is the load shedding power of petal network load node i at time t.

[0090] DSO takes the load balancing degree of the petal network and the minimum fault scheduling cost as the comprehensive goal, and optimizes the operation status of the petal network after the fault.

[0091]

[0092] Where: is the load balancing degree of the petal network in each discrete time period Δt; T is the total duration of fault repair.

[0093] The failure scheduling cost W of Huaban Network sch Including network loss cost W loss 、PIES incentive cost W PIES and IL compensation cost W cut , as shown below:

[0094] W sch =W loss +W PIES +W cut (19)

[0095]

[0096] Where: is the active power loss of line i→j at time t; α is the converted price of network loss.

[0097]

[0098] Where: is the additional power of the kth PIES at time t; K is the number of PIES participating in fault self-healing.

[0099] Under normal operating conditions, the cost of electricity purchase and sale in each period of the Petal Network is the product of the time-of-use electricity price and the PIES interactive power, which is a constant for the incentive cost and is therefore not reflected in the fault dispatch cost. The incentive price paid by DSOs, which is higher than the time-of-use electricity price, is the motivation for PIES aggregators to respond to incentive-based DR. Within a certain range, the higher the incentive price, the more additional power PIES will issue, and the less load shedding power the Petal Network will have.

[0100] In addition, the strength and efficiency of PIES power generation also depends on the PIES access location and the topology of the petal network. According to the connectivity between the PIES access load node and different substations, it is divided into different areas. Figure 1 As shown, PIES1-2 is connected to substation A, and PIES3 is connected to substation B, which are located in regions 1 and 2 respectively. The load rates of blocked lines in different regions may be different, the incentive electricity prices set by DSO are also different, and the additional power of PIES will also be different.

[0101] Considering multiple factors such as the importance of IL users, their willingness to cut power during the current period, and the maximum load shedding capacity, DSO pays IL users load shedding compensation fees at multiple time scales, as shown below:

[0102]

[0103] Where: is the load shedding loss of node i at time t, which is characterized by a quadratic function curve model and The relationship between; k1, k2 are load loss coefficients, k1 = 0.002, k2 = 3.2; γ i is the compensation coefficient of node i; is the power outage willingness of IL users at time t, ranging from [0,1], The smaller it is, the lower its willingness to cut off power. i , All are related to the load level.

[0104] Sub-objective function and W sch After normalization similar to formula (9), it is the main game profit p1:

[0105]

[0106] Where: ω is the blocking risk factor; the meanings of other variables are similar to those in formula (9).

[0107] ω is determined by the proportion of load power at time t to the peak load power during the day. When the proportion is ≥50%, the line load rate is generally high and needs to be increased. The weight of ω is set to sacrifice a certain degree of economy to improve power supply reliability, and ω is set to 0.5. When the proportion is less than 50%, the line load rate is generally low, and the incentive DR is dominated by economy, and ω is set to 0.

[0108] (2) Game constraints of the subject

[0109] 1) Line flow constraints

[0110] Same as Section 2.2. In addition, further constraints are imposed on the line load rate:

[0111]

[0112] 2) Voltage and current constraints

[0113] Same as Section 2.2.

[0114] 3) Incentive electricity price constraints

[0115] In order to ensure that the incentive electricity price is always within a reasonable range and effectively play its incentive role, it is necessary to constrain the incentive electricity price for each period:

[0116]

[0117] Where: is the time-of-use electricity price at time t.

[0118] 4) Load shedding constraints

[0119]

[0120] Where: is the load power of node i at time t; i is the maximum shelvable load capacity ratio of IL users at node i.

[0121] 3.2. Slave Player Model

[0122] (1) Game strategies and benefits

[0123] The PIES energy equipment output and energy trading plan formulated by the game strategy s2 for the PIES aggregator:

[0124]

[0125] Where: are the output, electricity purchase and sales power, and gas purchase volume of the kth PIES energy device i at time t. Trading power by plan With additional power composition.

[0126] The income from the game is the operating income of PIES, that is, the energy sales income of all PIES, W ex and operating cost C sch Difference:

[0127] p2=max(W ex -C sch ) (28)

[0128] PIES Energy Sales Revenue W exIncluding incentive benefits W PIES , planned transaction income and internal load energy supply income, as shown below:

[0129]

[0130] Where: are the energy sales price and power of the kth PIES internal load j at time t respectively.

[0131] PIES operating costs C sch Including gas purchase cost C buy 、Environmental protection cost C en , Operation and maintenance cost C op and supply reduction cost C loss , as shown below:

[0132]

[0133] Where: ν g The time-of-use unit price of natural gas; is the output of the CHP unit at time t; f CHP,θ (·) are the treatment cost and emission function of CHP for polluted gas θ respectively; N g is the number of pollutant gas types; N E is the number of PIES energy equipment; f op,i (·) is the operation and maintenance cost function of energy equipment i; are the compensation unit price and reduced power of the kth PIES internal load j at time t respectively.

[0134] (2) Constraints of the Slave Game

[0135] 1) Energy balance constraints

[0136] The power balance constraints of the PIES power subgrid are as follows:

[0137]

[0138] The power balance constraints for the PIES thermal subnetwork are as follows:

[0139]

[0140] The natural gas balance constraints of the PIES gas energy subgrid are as follows:

[0141]

[0142] Where: They are respectively the electricity, heat and gas load demands inside the kth PIES at time t.

[0143] 2) Energy equipment output constraints

[0144] Distributed generators (DG) such as PV and WT usually operate near the maximum power point under current natural conditions.

[0145]

[0146] Where: is the actual output power of DG at time t; Minimum power to maintain DG operation; is the maximum tracking power of DG at time t, which serves as the upper limit of DG output at that moment.

[0147] The charging and discharging process of ES, HS and GS devices is usually simplified to a 0-1 linear model, and the operating constraints (taking ES as an example) are as follows:

[0148]

[0149] Where: are the charge states of ES at time t and the initial time, respectively; They are ES charging and discharging power respectively; They are ES charging efficiency, discharging efficiency and self-discharging efficiency respectively; They are the lower and upper limits of ES state of charge respectively; They are the upper limits of ES charging and discharging power respectively; is the charge state of ES at the end of charging and discharging, in order to ensure that the capacity of ES is sufficient in the next scheduling stage, and Same; Δt ES is the ES charge and discharge time.

[0150] The output constraints of other energy equipment are as follows:

[0151]

[0152] Where: are the lower limits of the output of CHP, P2G and EB equipment at time t respectively;

[0153] They are the upper limits of the output of CHP, P2G and EB equipment respectively.

[0154] 3) Energy trading constraints

[0155]

[0156] Where: is the upper limit of the interaction power between the kth PIES and the petal network; is the kth PIES

[0157] Upper limit of gas purchase volume.

[0158] 3.3. Nash equilibrium point of master-slave game

[0159] To sum up, the comprehensive model of the master-slave game is:

[0160] Γ={X;s;p} (37)

[0161] In the formula: X = {1, 2} is the set of players; s = {s1, s2} is the set of game strategies; p = {p1, p2} is the set of game benefits.

[0162] Among them, p1=p1(u,v), p2=p2(u,v), the incentive electricity price decision is u∈s1, and the PIES additional power decision is v∈s2. If the DSO chooses the strategy u∈s1, the feedback set B(u) of the PIES aggregator is:

[0163]

[0164] DSO calls B(u), refers to the strategy w∈s2, and optimizes its own strategy to u * ∈s1, so that the following holds:

[0165]

[0166] Then, the PIES aggregator uses the policy u * ∈s1, choose its own optimal strategy v * ∈B(u * ). Set the strategy set s = {u * ,v *} is called the Nash equilibrium point of the master-slave game. p1(u,v)≥p1(u * ,v * ) and p2(u,v)≤p2(u * ,v * ). Nash equilibrium is a "binding compromise" between game participants. After reaching the Nash equilibrium, no player will easily change his or her strategy, otherwise the game benefits of both himself and the other party will decrease.

[0167] 3.4. Lower-level optimization model solution process

[0168] The present invention uses the PSO algorithm to initialize and update the incentive electricity price of PIES in each area, and uses the commercial solver CPLEX to solve the master-slave game model. The specific process is as follows:

[0169] 1) Input the operation information of Huaban Network and PIES, set the PSO algorithm parameters, randomly initialize the incentive electricity price that meets the constraints in different regions, and pass it to the PIES aggregator;

[0170] 2) The PIES aggregator substitutes this incentive electricity price as a known condition into the slave player model, calculates and retains the current revenue, and feeds back the PIES power increase plan to the DSO;

[0171] 3) The DSO substitutes the incentive electricity price and the additional power plan as known conditions into the subject player model, calculates and retains the current revenue;

[0172] 4) Update the current individual optimal solution and the global optimal solution;

[0173] 5) Determine whether the master-slave game has reached the maximum number of iterations. If so, consider the game to have reached the Nash equilibrium point, end the algorithm, and output the result; otherwise, update the incentive electricity price and continue the iteration, repeating steps 2)-4).

[0174] 4. Dynamic fault self-healing strategy process of Huaban Network based on two-layer optimization model

[0175] In order to avoid system power imbalance caused by load power or PIES interaction power fluctuation during the fault repair phase, the present invention divides the fault repair phase into multiple discrete time periods to dynamically adjust the fault self-healing decision. The action time of each stage is as follows: Figure 6 The process of the Petal Network fault self-healing strategy based on the two-layer optimization model is shown in Figure 7 Specifically, when a failure occurs in the petal network, the steps of the strategy are as follows:

[0176] Step 1: DMS collects the operating parameters and switch status of the petal network in real time, quickly locates and isolates the faulty components according to the fault information and operating data, and completes the load transfer. The specific measures are shown in Table 1.

[0177] Table 1 Common permanent fault isolation and load transfer measures in Huaban Network

[0178]

[0179] Step 2: DMS determines whether there is a blocked line in the petal network according to the line load rate information. If a fault blockage occurs, proceed to the next step; otherwise, proceed to step 7;

[0180] Step 3: DMS solves the upper optimization model, performs topology search based on the BFS algorithm, and quickly calculates and generates the optimal network reconstruction plan under the current load power and PIES interaction power by taking into account the objective function and constraints established in Section 2. It sends the switch state conversion instruction to the switch station and simultaneously uploads the network reconstruction plan to the DSO as the basis for solving the lower optimization model.

[0181] Step 4: If the upper layer optimization model has no solution, that is, the network reconstruction fails, go directly to step 6; otherwise, go to the next step;

[0182] Step 5: DMS determines whether the fault blocking is eliminated. If there is still a blocked line, it proceeds to the next step; otherwise, it proceeds to step 7;

[0183] Step 6: DSO solves the lower-level optimization model, conducts a master-slave game with PIES aggregators through incentive electricity prices, considers multi-dimensional factors to compensate for the load shedding losses of IL users, and sends the optimal load shedding plan to DMS, which executes the load shedding instructions; PIES aggregators formulate the optimal operation plan for PIES, perform energy management on related PIES, and optimize and adjust the interaction power between PIES and Petal Network;

[0184] Step 7: DMS determines whether a new fault occurs in the Petal Network. If so, it restarts the fault self-healing strategy; otherwise, DMS determines that the Petal Network can operate safely and stably during the current period.

[0185] Although the embodiments and drawings of the present invention are disclosed for illustrative purposes, those skilled in the art will appreciate that various substitutions, changes and modifications are possible without departing from the spirit and scope of the present invention and the appended claims. Therefore, the scope of the present invention is not limited to the contents disclosed in the embodiments and drawings.

Claims

1. A petal-type distribution network fault self-healing method based on a two-layer optimization model, characterized in that: The steps of the method are: S1. Analyze the mechanism of PIES access to the petal network fault blocking, weight the power transfer entropy by the inverse of the line load rate, and obtain the fault-weighted power transfer entropy of the petal network. Use the "entropy change" process of the power flow after the fault to quantify the probability of the petal network line blocking and the severity of the fault blocking, as the theoretical basis for solving the double-layer optimization model; S2. Considering the reliability and economy of the self-healing of the petal network fault, a two-layer optimization model based on "reconstruction-incentive" is established. The DMS uses the switch state of the petal network as the decision variable to solve the upper-layer optimization model based on network reconstruction, including the objective functions of load balancing, network loss and switch operation simplicity, as well as the constraints of network topology constraints, line flow constraints, and voltage and current constraints, as the basis for solving the lower-layer optimization model; S3, DSO uses the load shedding power of the petal network and the additional power of PIES as decision variables to solve the lower-level optimization model based on incentive-based DR. The main player is DSO, the game strategy is the incentive electricity price and the load shedding power of the petal network, the game benefit is the comprehensive objective function of the load balancing degree of the petal network and the fault dispatching cost, and the game constraints are line flow constraints, voltage and current constraints, incentive electricity price constraints and load shedding constraints; the slave player is the PIES aggregator, the game strategy is the PIES energy equipment output and energy trading plan, the game benefit is the PIES operating income, and the game constraints are energy balance constraints, energy equipment output constraints and energy trading constraints; finally, DSO and PIES aggregator realize game interaction and collaborative optimization through incentive electricity prices, among which DSO considers multi-dimensional factors to compensate for the load shedding losses of IL users, and PIES aggregator optimizes and adjusts the interactive power between PIES and petal network; S4. Design a dynamic fault self-healing strategy process for the petal network based on a two-layer optimization model, divide the petal network fault repair phase into multiple discrete time periods, and dynamically adjust the fault self-healing decision.

2. The petal-type distribution network fault self-healing method based on a double-layer optimization model according to claim 1 is characterized in that: S4 is specifically: when a failure occurs in the petal network, the strategy steps are as follows: 1) DMS collects the operating parameters and switch status of the petal network in real time, quickly locates and isolates the faulty components based on the fault information and operating data, and completes the load transfer; 2) DMS determines whether there is a blocked line in the petal network based on the line load rate information. If a fault blockage occurs, proceed to the next step; otherwise, go to 7); 3) DMS solves the upper optimization model, performs topology search based on the BFS algorithm, and quickly calculates and generates the optimal network reconstruction plan under the current load power and PIES interaction power by taking into account the objective function and constraints established by S2, sends the switch state conversion instruction to the switch station, and simultaneously uploads the network reconstruction plan to the DSO as the basis for solving the lower optimization model; 4) If the upper layer optimization model has no solution, that is, the network reconstruction fails, go directly to 6); otherwise, go to the next step; 5) DMS determines whether the fault blockage has been eliminated. If the blocked line still exists, it proceeds to the next step; Otherwise, go to 7); 6) DSO solves the lower-level optimization model, conducts a master-slave game with PIES aggregators through incentive electricity prices, considers multi-dimensional factors to compensate for the load shedding losses of IL users, and sends the optimal load shedding plan to DMS, which executes the load shedding instructions; PIES aggregators formulate the optimal operation plan of PIES, perform energy management on related PIES, and optimize and adjust the interaction power between PIES and Petal Network; 7) DMS determines whether a new fault occurs in the petal network. If so, it restarts the fault self-healing strategy; Otherwise, DMS determines that the Huaban Network is operating safely and stably during the current period.

Images