Power distribution network post-disaster maintenance method and system considering repair time sequence dependency

Abstract

The invention belongs to the technical field of power distribution network maintenance, and particularly relates to a power distribution network post-disaster maintenance method and system considering repair time sequence dependency, and the method comprises the steps: obtaining fault point information, traffic network information and base station information of a power distribution network after natural disasters, and calculating the generalized passing time of each road of a post-disaster traffic network; according to the obtained generalized passing time, determining a maintenance shortest path considering post-disaster road damage; according to the determined shortest maintenance path, calculating and considering the maintenance pre-time of post-disaster road damage, and meanwhile, calculating the maintenance duration by considering the fatigue degree of maintenance personnel and the real-time influence of extreme weather; in combination with the maintenance pre-time and the maintenance duration, taking the minimum total cost as a target, considering the maintenance sequence constraint, the element commissioning constraint and the power distribution network operation constraint, and constructing a power distribution network post-disaster maintenance model considering the repair time dependency; and solving the constructed post-disaster maintenance model of the power distribution network to obtain a post-disaster maintenance scheme of the power distribution network, and completing post-disaster maintenance of the power distribution network considering the repair time sequence dependency.

Description

Technical Field

[0001] The present invention belongs to the technical field of distribution network maintenance, and in particular relates to a distribution network post-disaster maintenance method and system taking into account the dependence of repair time series. Background Art

[0002] The statements in this section merely provide background information related to the present invention and do not necessarily constitute prior art.

[0003] The frequent occurrence of extreme disasters such as typhoons, rainstorms and waterlogging can lead to damage to power equipment, and even damage to power equipment with a large range and long duration; it is very easy to cause large-scale power outages, causing a chain reaction in society, and then bring huge economic losses and social impacts; at the same time, waterlogging on roads will also seriously hinder the dispatch of power restoration resources.

[0004] Through effective coordination in the post-disaster fault repair and recovery phase, the load can be quickly and stably restored. The focus is on improving the response speed and recovery capability of the power grid. The key issue is how to effectively regulate various emergency resources under the conditions of limited information and limited resources, optimize emergency recovery strategies, and restore load supply as soon as possible.

[0005] The existing post-disaster maintenance and recovery of distribution networks can be carried out around the coordination of multiple types of resources, optimization of emergency repair strategies, etc.: In terms of the coordination of multiple types of resources, the core of related research lies in the modeling method, that is, how to effectively model the operating characteristics and working mechanisms of various types of resources in order to coordinate and optimize them; however, the impact of repair time series dependency on the post-disaster maintenance and recovery of distribution networks is not considered. At the same time, road damage, fatigue of maintenance personnel and real-time impact of extreme weather are not fully considered in the post-disaster maintenance process. Summary of the invention

[0006] To solve the above problems, the present invention proposes a distribution network post-disaster maintenance method and system taking into account the dependence of repair time series, fully considering the impact of the dependence of repair time series on the post-disaster maintenance and recovery of the distribution network, and on the premise of meeting the constraints of safe and stable operation of the distribution network, with the goal of minimizing economic losses, generating a post-disaster maintenance plan for post-disaster maintenance of the distribution network, thereby improving the safety and reliability of the distribution network operation.

[0007] According to some embodiments, a first solution of the present invention provides a distribution network post-disaster maintenance method taking into account the dependency of repair time series, using the following technical solutions:

[0008] A distribution network post-disaster maintenance method taking into account the dependence of repair time series, comprising:

[0009] Obtain the fault point information, traffic network information and base station information of the distribution network after the natural disaster, and calculate the generalized travel time of each road in the post-disaster traffic network;

[0010] Based on the obtained generalized travel time, determine the shortest repair path considering the post-disaster road damage;

[0011] Based on the determined shortest repair path, calculate the repair lead time taking into account the post-disaster road damage. At the same time, consider the fatigue of the repair personnel and the real-time impact of extreme weather to calculate the repair duration.

[0012] Combining the obtained maintenance lead time and maintenance duration, taking the minimum total cost as the goal, considering the maintenance sequence constraints, component commissioning constraints and distribution network operation constraints, a distribution network post-disaster maintenance model taking into account the dependence of repair time is constructed;

[0013] Solve the constructed distribution network post-disaster maintenance model, obtain the distribution network post-disaster maintenance plan, and complete the distribution network post-disaster maintenance taking into account the dependence of the repair time series.

[0014] As a further technical limitation, the process of calculating the generalized travel time of each road in the post-disaster transportation network is: mapping the acquired fault point information of the distribution network after a natural disaster to the transportation network to obtain the post-disaster virtual network node of the distribution network; and calculating the generalized travel time of each road in the post-disaster transportation network based on the actual road length, road damage rate and road congestion coefficient in the transportation network information.

[0015] As a further technical limitation, the generalized travel time of each road in the post-disaster traffic network is used as the weight between virtual network nodes to generate an adjacency matrix of the traffic network with time weights; based on the generated adjacency matrix and the minimum path algorithm, the shortest repair path that takes into account the post-disaster road damage is obtained.

[0016] As a further technical limitation, the problem of severely damaged roads that maintenance teams cannot reach by car is taken into account, and the maintenance lead time is calculated based on drone technology and the establishment of emergency channels. The fatigue of maintenance personnel and the real-time impact of extreme weather are taken into account, and the maintenance duration is calculated based on the maintenance efficiency.

[0017] As a further technical limitation, the distribution network post-disaster maintenance model aims to minimize the system's cumulative load supply shortage and the latest completion time of all maintenance tasks, and comprehensively optimizes the movement path of maintenance personnel, the order of commissioning of repaired transmission lines, and the optimal power flow problem of the power grid.

[0018] As a further technical limitation, the distribution network post-disaster maintenance model constructed by the solution is divided into two stages: the first stage searches for a better candidate maintenance sequence through a heterogeneous dual-population ant colony algorithm based on a collaborative filtering strategy; the second stage optimizes the commissioning order of the transmission lines through a cyclic timing right shift strategy, improves the objective function corresponding to the maintenance sequence, and finally determines the optimal solution to obtain a distribution network post-disaster maintenance plan.

[0019] According to some embodiments, a second solution of the present invention provides a distribution network post-disaster maintenance system taking into account the dependency of repair time series, which adopts the following technical solution:

[0020] A distribution network post-disaster maintenance system taking into account the dependence of repair time series, comprising:

[0021] An acquisition module, which is configured to acquire fault point information, traffic network information, and base station information of the distribution network after a natural disaster, and calculate the generalized travel time of each road in the traffic network after the disaster;

[0022] A determination module, which is configured to determine the shortest repair path considering the post-disaster road damage according to the obtained generalized travel time;

[0023] A calculation module is configured to calculate the maintenance lead time considering the post-disaster road damage according to the determined shortest maintenance path, and at the same time, calculate the maintenance duration considering the fatigue of maintenance personnel and the real-time impact of extreme weather;

[0024] A modeling module is configured to combine the obtained maintenance lead time and maintenance duration, take the minimum total cost as the goal, consider the maintenance sequence constraints, component commissioning constraints and distribution network operation constraints, and construct a distribution network post-disaster maintenance model that takes into account the dependence of repair time;

[0025] The maintenance module is configured to solve the constructed distribution network post-disaster maintenance model, obtain the distribution network post-disaster maintenance plan, and complete the distribution network post-disaster maintenance taking into account the dependence of the repair time series.

[0026] According to some embodiments, a third solution of the present invention provides a computer-readable storage medium, which adopts the following technical solution:

[0027] A computer-readable storage medium stores a program, which, when executed by a processor, implements the steps of a distribution network post-disaster maintenance method taking into account the dependence of repair time series as described in the first solution of the present invention.

[0028] According to some embodiments, a fourth solution of the present invention provides an electronic device, which adopts the following technical solution:

[0029] An electronic device comprises a memory, a processor and a program stored in the memory and running on the processor, wherein when the processor executes the program, the steps of the distribution network post-disaster maintenance method taking into account the dependence of the repair time series as described in the first scheme of the present invention are implemented.

[0030] According to some embodiments, a fifth solution of the present invention provides a computer program product, which adopts the following technical solution:

[0031] A computer program product includes software codes, wherein the program in the software codes executes the steps of the distribution network post-disaster maintenance method taking into account the dependence of repair time series as described in the first embodiment of the present invention.

[0032] According to some embodiments, the sixth scheme of the present invention provides specific embodiments to illustrate the effects of the present invention.

[0033] Compared with the prior art, the present invention has the following beneficial effects:

[0034] The present invention fully considers the dependence of the repair time series when performing post-disaster maintenance of the distribution network, and constructs a repair time model with sequence dependence, including a maintenance lead time model considering the impact of road damage and a maintenance duration model considering the fatigue of maintenance personnel and the real-time impact of extreme weather; under the premise of satisfying the maintenance sequence constraints, component commissioning constraints and system operation constraints, the maintenance team, emergency repair materials, etc. are coordinated to minimize the system cumulative load supply shortage and the latest completion time of all maintenance tasks, and the distribution network post-disaster maintenance model is constructed, and the movement path of the maintenance personnel, the commissioning order of the repaired transmission lines and the optimal power flow problem of the power grid are comprehensively optimized; and the constructed distribution network post-disaster maintenance model is then solved in stages: that is, in the first stage, a better candidate maintenance sequence is searched through a heterogeneous dual-population ant colony algorithm based on a collaborative filtering strategy, and in the second stage, the commissioning order of the transmission lines is optimized through a cyclic timing right shift strategy, the objective function corresponding to the maintenance sequence is improved, and finally the optimal solution is determined to obtain a distribution network post-disaster maintenance plan. The present invention fully considers that the dependence of repair time series will have a great impact on maintenance decisions. Under the premise of meeting the constraints of safe and stable operation of the distribution network, the goal of minimizing economic losses is achieved and a post-disaster maintenance strategy is generated. BRIEF DESCRIPTION OF THE DRAWINGS

[0035] The drawings in the specification that constitute a part of this embodiment are used to provide a further understanding of this embodiment. The schematic embodiments of this embodiment and their descriptions are used to explain this embodiment and do not constitute improper limitations on this embodiment.

[0036] Figure 1 It is a flow chart of a distribution network post-disaster maintenance method taking into account the dependence of repair time series in the first embodiment of the present invention;

[0037] Figure 2 A Gantt chart corresponding to the maintenance sequence in the first embodiment of the present invention;

[0038] Figure 3 This is a flow chart of a two-stage distribution optimization solution algorithm in Embodiment 1 of the present invention;

[0039] Figure 4It is a structural block diagram of a distribution network post-disaster maintenance system taking into account the dependency of repair time series in the second embodiment of the present invention;

[0040] Figure 5 The IEEE-30 node system used in the sixth embodiment of the present invention;

[0041] Figure 6 is the load supply level recovery curve in the sixth embodiment of the present invention;

[0042] Figure 7 The IEEE-118 node system used in the seventh embodiment of the present invention;

[0043] Figure 8 This is the load supply level recovery curve in the seventh embodiment of the present invention. DETAILED DESCRIPTION

[0044] The present invention will be further described below in conjunction with the accompanying drawings and embodiments.

[0045] It should be noted that the following detailed descriptions are exemplary and are intended to provide further explanation of the present application. Unless otherwise specified, all technical and scientific terms used herein have the same meanings as those commonly understood by those skilled in the art to which the present application belongs.

[0046] It should be noted that the terms used herein are only for describing specific embodiments and are not intended to limit exemplary embodiments according to the present invention. As used herein, unless the context clearly indicates otherwise, the singular form is also intended to include the plural form. In addition, it should be understood that when the terms "comprising" and / or "including" are used in this specification, it indicates the presence of features, steps, operations, devices, components and / or combinations thereof.

[0047] In the absence of conflict, the embodiments of the present invention and the features of the embodiments may be combined with each other.

[0048] Embodiment 1

[0049] Embodiment 1 of the present invention introduces a distribution network post-disaster maintenance method taking into account the dependence of repair time series.

[0050] like Figure 1 A distribution network post-disaster maintenance method taking into account the dependence of repair time series is shown, comprising the following steps:

[0051] Step S01: Collect non-electric information of road congestion or interruption under extreme disasters, map the fault point in the distribution network to the transportation network, and form a virtual network node with the emergency center in the transportation network;

[0052] Step S02: Based on the obtained information, the generalized travel time of each road after the disaster is obtained, and the travel time is used as the weight between virtual network nodes to generate an adjacency matrix of the traffic network with time weights;

[0053] Step S03: using Dijkstra's minimum path algorithm to search for the shortest path between maintenance tasks or between a maintenance task and a base station;

[0054] Step S04: Considering the problem that the maintenance team cannot reach the severely damaged road area by car, the maintenance lead time is calculated based on drone technology and the establishment of emergency channels;

[0055] Step S05: Considering the real-time impact of maintenance personnel fatigue and extreme weather, the maintenance duration is calculated based on the maintenance efficiency;

[0056] Step S06: Under the premise of satisfying the maintenance sequence constraints, component commissioning constraints and system operation constraints of the distribution network maintenance decision model, a post-disaster maintenance decision model taking into account the dependence of the repair time series is established;

[0057] Step S07: Separate the original model into two problems: maintenance sequence search and line commissioning sequence optimization. Solve the first stage through the heterogeneous dual-population ant colony algorithm based on collaborative filtering strategy, and solve the second stage through the cyclic timing right shift strategy. Improve the objective function corresponding to the maintenance sequence and finally determine the optimal solution.

[0058] As one or more implementation modes, in step S01, the information of the fault point in the distribution network after the disaster, the damage of the fault point, the degree of road damage, the road congestion coefficient and other data are obtained through the big data platform, and the location of the base station and the information of the maintenance team are obtained from the power department.

[0059] As one or more implementation modes, in step S02, by introducing the generalized road length d ij To measure the impact of road damage on the maintenance team's travel time:

[0060]

[0061] Among them, d ij ′Actual road length; θ ij is the road damage rate; δ is the threshold; M is an infinite number, indicating that when θ ij When it is greater than or equal to the threshold, the road is impassable.

[0062] The congestion of the road is reflected in the maintenance team's driving speed v ij Impact:

[0063]

[0064] in, is the road congestion coefficient;

[0065] That is, the generalized travel time of the post-disaster maintenance team on road ij is:

[0066]

[0067] The generalized travel time of each road after the disaster is ti j As the weight between virtual network nodes, to generate the adjacency matrix A of the traffic network with time weight t .

[0068] As one or more implementation modes, in step S03, when the starting point S of the maintenance team is known at this time, t and the end point F t When , the Dijkstra minimum path algorithm can be called to search for the shortest path from the current location of the maintenance team to the target maintenance task or base station at time t;

[0069] D1{·}=Dijkstra(A t ,S t ,F t );

[0070] D2{·}=Dijkstra(1,S t ,F t );

[0071] The function D1{·} represents the shortest path length between multiple geographical locations considering post-disaster road damage; and D2{·} represents the shortest path length without considering post-disaster road damage.

[0072] like Figure 2 As shown, the dependency of component repair time on the maintenance sequence is illustrated.

[0073] As one or more implementation methods, in step S04, considering the problem that the maintenance team cannot drive to the severely damaged road area, the present invention uses a post-disaster repair material dispatching strategy of the maintenance team-UAV collaboration and the establishment of emergency traffic channels to ensure the repair of important faulty components in the restricted area. Therefore, the maintenance lead time is divided into 3 parts:

[0074] (1) Lead time for repair of faulty components within the area accessible by the maintenance team:

[0075] For the area accessible to the maintenance team, consider the material replenishment process, that is,

[0076]

[0077] (2) Lead time for repair of faulty components in areas with restricted access by maintenance teams:

[0078] For the area where the maintenance team has restricted access, it is assumed that the maintenance materials in the original material vehicle of the maintenance team can meet the needs of the faulty components in the restricted area; that is,

[0079]

[0080] (3) The lead time for repair of faulty components at the junction of the restricted access area and the accessible area for maintenance teams:

[0081]

[0082] in, and F i They represent the remaining maintenance materials of maintenance team r after completing task i and the maintenance materials required for task i respectively; d(i′,i) represents the serial number of the base station for material replenishment between tasks i and i′; represents the travel speed of the maintenance team r; It means that the replenishment of materials takes time; Indicates the travel speed of the drone; Indicates the speed at which maintenance personnel travel on the emergency traffic lane.

[0083] As one or more implementation modes, in step S05, it is assumed that a maintenance task i starts at time s i , the maintenance duration of the task is obtained by solving the following equation:

[0084]

[0085] Among them, v r,i,t (W t ),v r,i,t (μ t ) represent the weather intensity W t and failure degree μ t The standardized maintenance efficiency of maintenance team r for maintenance task i; α represents the weight ratio of the two efficiencies.

[0086] R r,i (s i ) can be regarded as the maintenance start time s i function, and s i Determined by the maintenance sequence. The number 1 on the right side of the above formula represents the total workload of the standardized maintenance task.

[0087] This embodiment takes typhoon extreme weather as an example, and the maintenance efficiency v r,i,t (W t ) takes the following expression:

[0088]

[0089] Among them, W t The ten-minute average wind speed at the surface is used to characterize meteorological intensity; represents the average maintenance duration under normal working conditions; α r,i and β r,i Function parameter representing the variation of maintenance efficiency with wind speed.

[0090] The Weibull distribution is used to measure the failure rate of the maintenance team, thus obtaining the maintenance efficiency v affected by the physical strength of the maintenance personnel. r,i,t (μ t ) expression:

[0091]

[0092]

[0093] Among them, μ t is the failure degree; k is the shape parameter, λ is the scaling factor;

[0094] The above repair time model is used to reflect the maintenance sequence and maintenance lead time L through constraints. r,i , Maintenance duration R r,i (s i ) between them.

[0095] As one or more implementation methods, in step S06, the optimization goal of the distribution network post-disaster maintenance model is to minimize the system cumulative load supply shortage and the latest completion time of all maintenance tasks. The power system restoration considering component maintenance is divided into the following two stages:

[0096] Phase 1: The main goal of this phase is to reduce the load supply shortfall as much as possible in a shorter period of time. In other words, the cumulative load supply shortfall should be minimized.

[0097] The objective function of stage one is:

[0098]

[0099] in, represents the load supply value of node b, that is, the economic loss per unit of load loss, and They represent the node load supply shortage ratio and the node ideal load supply level respectively; T is the maximum duration of the recovery phase, and the single time period is generally set to 1 hour.

[0100] Phase 2: When the system reaches the ideal load supply level, there may still be redundant components that have not been repaired. At this time, the maintenance decision problem is transformed into a traditional path-finding problem. The dispatcher needs to minimize the latest completion time of the maintenance process, that is, the end time of the last maintenance task:

[0101] Ψ2 = minχ;

[0102] Where χ represents the latest completion time of all maintenance teams.

[0103] Combining the objective functions of the two stages, the overall objective function of the maintenance decision problem is:

[0104]

[0105] Among them, ω1 and ω2 represent the weights of the objective functions of stage one and stage two, respectively.

[0106] The maintenance sequence constraints, component commissioning constraints, and system operation constraints of the distribution network post-disaster maintenance decision model are specifically:

[0107] (1) Maintenance sequence constraints:

[0108] Each maintenance task node must be preceded and followed by another task node / starting node / ending node, and only one maintenance team is allowed to access it:

[0109]

[0110] Among them, X r,i,i″ Used to represent the maintenance sequence. If the next task of maintenance team r after completing maintenance task i is i″, then the binary integer variable X r,i,i″ =1, otherwise X r,i,i″ =0;Ω r represents the set of all maintenance teams; Ω a ,Ω d and Ω e They correspond to the task node, the start node and the set of virtual end nodes respectively. The virtual end node indicates the end of the maintenance sequence of all maintenance teams;

[0111] The maintenance team must start from the base station where they are stationed:

[0112]

[0113] Among them, ξ r,d is a binary parameter, ξ r,d =1 means that the maintenance team r is located at base station d at the beginning;

[0114] If i is after i′ in the maintenance sequence, its maintenance start time is equal to the end time of task i′ plus the lead time of task i:

[0115]

[0116] Among them, s i and f i They represent the start and end time of the repair of the faulty component i, respectively, and M is a sufficiently large constant;

[0117] The difference between the end and start time of the same maintenance task i is equal to the maintenance duration:

[0118]

[0119] The time at which each maintenance team is at the starting position is 0:

[0120]

[0121] The relationship between the end time of the maintenance task and the operational status of the faulty component:

[0122]

[0123] Among them, v i,t is the operational status of faulty component i in period t, v i,t =1 means that the system is allowed to be put into operation. Here, the constant M is used to round off the number of time periods.

[0124] The latest completion time of the entire maintenance process shall not be earlier than the end time of any maintenance task:

[0125]

[0126] In the above constraints, the maintenance lead time L r,i and the maintenance duration R r,i As a variable.

[0127] 1) Maintenance lead time L r,i :

[0128] Maintenance lead time L r,i Use the following constraints for specification:

[0129] Judge L r,i Whether the relationship defined in step S04 is satisfied:

[0130]

[0131] Among them, (·) are three situations in step S04;

[0132] Update the remaining supplies of the maintenance team:

[0133]

[0134] in, Indicates the maximum material carrying capacity of maintenance team r;

[0135] Set the initial amount of supplies the maintenance team can carry:

[0136]

[0137] The drone can take off or land at any node at most once:

[0138]

[0139] When the drone on the maintenance team r is launched at the faulty component i′, delivers supplies to the faulty component i, and is retrieved at the faulty component i″, is 1, otherwise it is 0;

[0140] The drone can only take off and land at two nodes that the maintenance team passes through consecutively, with the take-off node in front and the landing node in the back. At the same time, the drone is limited to delivering supplies to only one faulty component each time it is dispatched:

[0141]

[0142] The amount of supplies carried by the drone each time it is dispatched is less than or equal to its maximum load limit:

[0143]

[0144] When the materials of the faulty component i are delivered by the drone carried by the maintenance team r, is 1, otherwise it is 0; The maximum payload of the drone;

[0145] The mileage that the drone needs to travel each time is less than or equal to its range limit:

[0146]

[0147] Among them, Q D The range of a drone in a single sortie;

[0148] 2) Maintenance duration R r,i :

[0149] R r,i Itself and s i There is a nonlinear relationship. This problem is solved by i The discretization of s is solved. First, uniformly divide s in the interval [0,T] i Take K segmentation points. is the kth segmentation point, For the corresponding The maintenance duration can be calculated in advance for each segmentation point. The more segmentation points are set, the higher the approximation accuracy. Then, add the following constraints:

[0150] s i With its selected segment point The distance between them must be smaller than that between other segment points:

[0151]

[0152]

[0153]

[0154]

[0155] Among them, the binary variable o r,i,k =1 indicates distance s i The nearest segmentation point is By o r,i,k Will s i Approximately And use the variable Δs i,k Alternative Using an auxiliary binary variable w i,k express The positive and negative nature of

[0156] Only when maintenance team r performs task i, o r,i,k Only then can it be possible to take 1, and at most only a single segmentation point can be selected:

[0157]

[0158] Establish the maintenance duration R r,i and its approximate value Relationship:

[0159]

[0160] (2) Component commissioning constraints:

[0161] Once a substation at a node is operational, the substation can resume power transmission at any time:

[0162]

[0163] Among them, Ω b represents the set of node substations, b represents Ω b The index of the element in ;

[0164] When a transmission line has been repaired and both substations at both ends have been put into operation, the transmission line can be put into operation:

[0165]

[0166] Among them, Ω l represents the set of node substations, l represents Ω l The element index in ; s(l) and e(l) represent the first and last nodes of line l;

[0167] Once a component is put into operation, it will not be taken out of operation again:

[0168]

[0169] Among them, u i,t Represents the actual operating status of component i in time period t.

[0170] (3) System operation constraints:

[0171] Node active and reactive balance equations:

[0172]

[0173] in, and Respectively represent the active / reactive unit output / line power flow; V b,t Indicates the node voltage amplitude; and denote the node shunt conductance and susceptance respectively;

[0174] If the transmission line has been put into operation, the power flow of the line must satisfy the power flow equation:

[0175]

[0176] Among them, θ b,t represents the node voltage phase angle; and represent branch conductance and susceptance respectively.

[0177] The active and reactive output range constraints of the unit. If the substation where the unit is located is not put into operation, the unit output is 0:

[0178]

[0179] Limit the flow on each line:

[0180]

[0181] Specify the range of node voltage amplitude and phase angle:

[0182]

[0183] The load supply on each node cannot exceed its ideal load supply level:

[0184]

[0185] As one or more implementation modes, in step S07, Figure 3 As shown, the original model is separated into two problems: maintenance sequence search and line commissioning sequence optimization.

[0186] In the first stage, a heterogeneous dual-population ant colony algorithm based on collaborative filtering strategy is used to search for a better candidate maintenance sequence as an elite solution set; specifically:

[0187] Step S701: Initialize task relevance pheromone τ i′i and task location pheromone τ ip , for the MMAS algorithm, during the search process, τ i′i and τ ip The value range of is limited to [0.5,5.0]. All pheromones are initialized to 0.5 and are enhanced during the search process;

[0188] Step S702: construct two identical populations, based on the ACS and MMAS algorithms respectively, and construct the corresponding maintenance sequences through the following steps:

[0189] (1) Set the set of faulty components that are not scheduled for repair Ω u is the set of all faulty components, that is, Ω u ←Ω a The current task position is set to p←1, and the completion time of each maintenance team is χ r ←0;

[0190] (2) Select the maintenance team r with the earliest completion time * Insert the next maintenance task, insert r * The probability distribution of the maintenance sequence is as follows:

[0191] 1) Probability distribution of maintenance sequence based on ACS:

[0192]

[0193] 2) Probability distribution of maintenance sequence based on MMAS:

[0194]

[0195] Among them, π i represents the probability of task i being inserted into the maintenance sequence in the current round; i′ is the maintenance team r* The last task executed before; coefficient and Indicates the importance of the two pheromones;

[0196] (3) Update the maintenance team * The latest completion time, assuming that the currently inserted task is i * , then r * The latest completion time is updated to:

[0197]

[0198] At the same time, update i * Maintenance end time Since the maintenance sequence before the current task is known, the insertion task and It is given directly by the repair time model of sequence dependence;

[0199] (4) If Update p←p+1,Ω u ←Ω u / i * , return to (2); otherwise, end the process.

[0200] Step S703: The system topology changes only when the components are put into operation. Therefore, the entire recovery phase will have at most |Ω a |+1 topology. List the operating status of each component in the nth topology:

[0201]

[0202] satisfy:

[0203]

[0204] Among them, f (n) Indicates the maintenance completion time corresponding to the component at the nth position;

[0205] use Evaluate the objective function of the original model under different topologies for the corresponding maintenance sequence:

[0206]

[0207] At the same time, it also includes constraints (1), (2), and (3) in step S06;

[0208] The time period number t of the above objective function and constraint conditions is replaced by the topological number n;

[0209] Use the above method to evaluate each candidate maintenance sequence, and retain a set of maintenance sequences with the most ideal objective function values ​​into the elite solution set for pheromone update and second-stage optimization;

[0210] Step S704: Update pheromone:

[0211] (1) Pheromone update strategy based on ACS

[0212] 1) Local pheromone update:

[0213]

[0214] in, represents the local pheromone attenuation coefficient; τ0 represents the initial pheromone amount;

[0215] 2) Global pheromone update:

[0216]

[0217]

[0218] Where η represents the pheromone attenuation coefficient; C(i′,i) represents the number of times task i is placed after i′ in the maintenance sequence of all elite solutions, and C(i,p) represents the number of times task i is placed in position p; λ 1 and Represents the task relevance pheromone τ i′i Pheromone τ with task location ip The magnitude of the enhancement at each iteration;

[0219] (2) Pheromone update strategy based on MMAS:

[0220]

[0221]

[0222] Step S705: Introduce collaborative filtering strategy to search for the paths preferred by the optimal solution ants in both ACS and MMAS, and give pheromone rewards to the paths of ACS every ω iterations. The role of collaboration is to make the optimal solution ants drive the non-optimal ants to move closer to the preferred paths when constructing the paths, making the algorithm more directional; the role of filtering is to remove some relatively poor paths and speed up the convergence of the algorithm.

[0223] set up is the position matrix of all ants when the ACS ant colony algorithm is iterated T times (T is an integer multiple of ω), is the position matrix of all ants at the Tth iteration of the MMAS ant colony algorithm.

[0224]

[0225] Among them, m is the number of ants.

[0226] Assume that k and h are the ants that find the optimal solution in the Tth iteration of the ACS and MMAS algorithms, respectively, so the common preference path is:

[0227]

[0228] Among them, S (ACS,MMAS) The three consecutive identical elements shared by the two populations are saved and set as the common preference path; Step S706: reward the common preference path, additional pheromone reward factor:

[0229]

[0230] Among them, iter is the number of iterations during communication;

[0231] Therefore, the pheromone update for the public preference path is:

[0232]

[0233] Step S707: Adaptively adjust the AC frequency, assuming that the exchange probability P is DPACO Determine whether two populations exchange information; i.e.

[0234]

[0235] Among them, T is the number of iterations, and k is the optimal solution ant.

[0236] When the exchange probability is greater than or equal to 0.5, it indicates that the diversity of the dual population algorithm has decreased, so there will be no communication between populations, otherwise communication will be carried out.

[0237] Therefore, the AC frequency of the entire DPACO algorithm is:

[0238]

[0239] Step S708: Exchange and averaging of pheromones. When the two algorithms stagnate, we exchange the pheromone tables of ACS and MMAS, thereby effectively jumping out of the local optimum;

[0240]

[0241] Step S709: The search process is terminated after the following criteria are met, specifically:

[0242] (1) The maximum number of iterations or computing time limit is reached;

[0243] (2) The average objective function value of the elite solution set changes less than a small threshold.

[0244] Phase II: Based on the results of the first phase, the operation sequence of the transmission lines is optimized through the cyclic timing right shift strategy, the objective function corresponding to the maintenance sequence is improved, and the optimal solution is finally determined; that is,

[0245] Step S710: Keep the commissioning time of other components unchanged, postpone the commissioning time of a line (move the commissioning time to the right on the time axis) to the commissioning time of other components, until the objective function cannot be further reduced, and any repaired component has only a finite number of possible commissioning times:

[0246]

[0247] Wherein, j1,...,jk,... represent the operation time of all components that are put into operation later than component i, and are arranged in ascending order according to the operation time; specifically:

[0248] (1) Select any maintenance sequence from the elite solution set in the first stage to enter the second stage optimization, and select any fault line i;

[0249] (2) Keep the commissioning time of other components unchanged and postpone the commissioning time of the selected line to Re-evaluate the objective function value. If the objective function value decreases, further postpone to And so on, until the objective function cannot be improved;

[0250] (3) Select the next line and repeat the operation (2). When all lines have been traversed, start the loop traversal again from the first line. When all lines have been traversed and the objective function has not been improved within a certain number of iterations, the optimization of the commissioning order of the maintenance sequence is terminated;

[0251] (4) Perform the above operations on all maintenance sequences in the elite solution set. The elite solution with the smallest objective function after the second stage optimization is taken as the optimal maintenance sequence.

[0252] By repeating this operation for all transmission lines, the rapid commissioning order of the maintenance sequence can be optimized, and finally the optimal solution can be screened out.

[0253] Embodiment 1 of the present invention fully considers the dependence of the repair time series when performing post-disaster maintenance of the distribution network, and constructs a repair time model with sequence dependence, including a maintenance lead time model considering the impact of road damage and a maintenance duration model considering the fatigue of maintenance personnel and the real-time impact of extreme weather; under the premise of satisfying the maintenance sequence constraints, component commissioning constraints and system operation constraints, coordinates maintenance teams, emergency repair materials, etc., with the goal of minimizing the system's cumulative load supply shortage and the latest completion time of all maintenance tasks, constructs a distribution network post-disaster maintenance model, and comprehensively optimizes the movement path of maintenance personnel, the commissioning order of repaired transmission lines, and the optimal power flow problem of the power grid; and then solves the constructed post-disaster maintenance model in stages: that is, in the first stage, a better candidate maintenance sequence is searched through a heterogeneous dual-population ant colony algorithm based on a collaborative filtering strategy, and in the second stage, the commissioning order of transmission lines is optimized through a cyclic timing right shift strategy, the objective function corresponding to the maintenance sequence is improved, and finally the optimal solution is determined to obtain a distribution network post-disaster maintenance plan. The present invention fully considers that the dependence of repair time series will have a great impact on maintenance decisions. Under the premise of meeting the constraints of safe and stable operation of the distribution network, the goal of minimizing economic losses is achieved and a post-disaster maintenance strategy is generated.

[0254] Embodiment 2

[0255] The second embodiment of the present invention introduces a distribution network post-disaster maintenance system that takes into account the dependence of repair time series.

[0256] like Figure 4 A distribution network post-disaster maintenance system taking into account the dependence of repair time series is shown, comprising:

[0257] An acquisition module, which is configured to acquire fault point information, traffic network information, and base station information of the distribution network after a natural disaster, and calculate the generalized travel time of each road in the traffic network after the disaster;

[0258] A determination module, which is configured to determine the shortest repair path considering the post-disaster road damage according to the obtained generalized travel time;

[0259] A calculation module is configured to calculate the maintenance lead time considering the post-disaster road damage according to the determined shortest maintenance path, and at the same time, calculate the maintenance duration considering the fatigue of maintenance personnel and the real-time impact of extreme weather;

[0260] A modeling module is configured to combine the obtained maintenance lead time and maintenance duration, take the minimum total cost as the goal, consider the maintenance sequence constraints, component commissioning constraints and distribution network operation constraints, and construct a distribution network post-disaster maintenance model that takes into account the dependence of repair time;

[0261] The maintenance module is configured to solve the constructed distribution network post-disaster maintenance model, obtain the distribution network post-disaster maintenance plan, and complete the distribution network post-disaster maintenance taking into account the dependence of the repair time series.

[0262] The detailed steps are the same as those of the distribution network post-disaster maintenance method taking into account the dependence of the repair time series provided in the first embodiment, and will not be repeated here.

[0263] Embodiment 3

[0264] Embodiment 3 of the present invention provides a computer-readable storage medium.

[0265] A computer-readable storage medium stores a program, which, when executed by a processor, implements the steps of the distribution network post-disaster maintenance method taking into account the dependence of repair time series as described in the first embodiment of the present invention.

[0266] The detailed steps are the same as those of the distribution network post-disaster maintenance method taking into account the dependence of the repair time series provided in the first embodiment, and will not be repeated here.

[0267] Embodiment 4

[0268] A fourth embodiment of the present invention provides an electronic device.

[0269] An electronic device comprises a memory, a processor and a program stored in the memory and running on the processor, wherein when the processor executes the program, the steps of the distribution network post-disaster maintenance method taking into account the dependence of the repair time series as described in the first embodiment of the present invention are implemented.

[0270] The detailed steps are the same as those of the distribution network post-disaster maintenance method taking into account the dependence of the repair time series provided in the first embodiment, and will not be repeated here.

[0271] Embodiment 5

[0272] Embodiment 5 of the present invention provides a computer program product.

[0273] A computer program product includes software codes, wherein the program in the software codes executes the steps in the distribution network post-disaster maintenance method taking into account the dependence of repair time series as described in the first embodiment of the present invention.

[0274] The detailed steps are the same as those of the distribution network post-disaster maintenance method taking into account the dependence of the repair time series provided in the first embodiment, and will not be repeated here.

[0275] Embodiment 6

[0276] The sixth embodiment of the present invention adopts the IEEE-30 node system mainly to verify the efficiency of different solution methods. Its topology is as follows: Figure 5 As shown;

[0277] The following five methods are set up for comparison:

[0278] (1) M1: Use a two-stage step-by-step algorithm to solve the maintenance decision model;

[0279] (2) M2: directly use the solver to solve the maintenance decision model, the optimality gap is set to 50%, and the maintenance area constraint is not included;

[0280] (3) M3: The optimality gap is set to 30%, and other settings are the same as M2;

[0281] (4) M4: The optimality gap is set to 10%, and other settings are the same as M2;

[0282] (5) M5: Add maintenance area constraints, set the optimality gap to 0.1%, and other settings are the same as M2.

[0283] Mathematical programming solutions M2-M4 cannot obtain the global optimal solution within 24 hours, so a larger optimality gap needs to be set as the convergence threshold. The lower the optimality gap, the higher the quality of the maintenance sequence obtained.

[0284] As can be seen from Table 1, as the optimality gap decreases, the solution time of the mathematical programming solution increases exponentially. The proposed two-stage optimization algorithm can obtain a better maintenance sequence in a short time. For example, the post-disaster losses generated by M1 and M4 are relatively similar, but the calculation time of the former is only 1.2% of the latter. Obviously, only the proposed algorithm meets the needs of rapid post-disaster response.

[0285] Table 1 Comparison of results under different methods

[0286]

[0287] The load supply level recovery curve corresponding to M1-M5 is as follows Figure 6 The cumulative load supply deficit (area above the curve) corresponding to M1 is the smallest, and the time to reach the ideal load supply level (horizontal dotted line with ordinate of 0) is second only to M4.

[0288] Embodiment 7

[0289] Embodiment 7 of the present invention uses the IEEE-118 node system to analyze the influence of the repair time series dependency on the result, and its topology is as follows: Figure 7 As shown;

[0290] The following scenario is set up to demonstrate the importance of considering fixing time series dependencies:

[0291] (1) Scenario 1: The maintenance duration is set to the average maintenance duration under normal weather conditions, and the lead time only considers the travel time of the maintenance team between components;

[0292] (2) Scenario 2: Considering the time-varying maintenance duration under the influence of weather conditions, the lead time also considers the travel time of the maintenance team between components and the material replenishment process;

[0293] Figure 8 The load supply level recovery curves corresponding to Scenario 1 and Scenario 2 are given. In both scenarios, the system can recover 75% of the load supply deficit within 25 hours. However, the recovery speed of Scenario 2 is significantly faster than that of Scenario 1, which reflects the importance of considering the dependence of the repair time series from the perspective of system operation.

[0294] The above description is only a preferred embodiment of the present embodiment and is not intended to limit the present embodiment. For those skilled in the art, the present embodiment may have various modifications and variations. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present embodiment shall be included in the protection scope of the present embodiment.

Claims

1. A distribution network post-disaster maintenance method taking into account the dependence of repair time series, characterized in that: include: Obtain the fault point information, traffic network information and base station information of the distribution network after the natural disaster, and calculate the generalized travel time of each road in the post-disaster traffic network; Based on the obtained generalized travel time, determine the shortest repair path considering the post-disaster road damage; Based on the determined shortest repair path, calculate the repair lead time taking into account the post-disaster road damage. At the same time, consider the fatigue of the repair personnel and the real-time impact of extreme weather to calculate the repair duration. Combining the obtained maintenance lead time and maintenance duration, taking the minimum total cost as the goal, considering the maintenance sequence constraints, component commissioning constraints and distribution network operation constraints, a distribution network post-disaster maintenance model taking into account the dependence of repair time is constructed; Solve the constructed distribution network post-disaster maintenance model, obtain the distribution network post-disaster maintenance plan, and complete the distribution network post-disaster maintenance taking into account the dependence of the repair time series.

2. A distribution network post-disaster maintenance method taking into account the dependence of repair time series as described in claim 1, characterized in that: The process of calculating the generalized travel time of each road in the post-disaster transportation network is as follows: mapping the acquired fault point information of the distribution network after a natural disaster to the transportation network to obtain the post-disaster virtual network node of the distribution network; and calculating the generalized travel time of each road in the post-disaster transportation network based on the actual road length, road damage rate and road congestion coefficient in the transportation network information.

3. A method for post-disaster maintenance of a distribution network taking into account the dependence of repair time series as claimed in claim 1, characterized in that: The generalized travel time of each road in the post-disaster traffic network is used as the weight between virtual network nodes to generate an adjacency matrix of the traffic network with time weights. Based on the generated adjacency matrix and the minimum path algorithm, the shortest repair path considering the post-disaster road damage is obtained.

4. A method for post-disaster maintenance of a distribution network taking into account the dependence of repair time series as claimed in claim 1, characterized in that: The problem of severely damaged roads that maintenance teams cannot reach by car is taken into consideration, and the maintenance lead time is calculated based on drone technology and the establishment of emergency channels. The maintenance duration is calculated based on the maintenance efficiency, taking into account the fatigue of maintenance personnel and the real-time impact of extreme weather.

5. A method for post-disaster maintenance of a distribution network taking into account the dependence of repair time series as claimed in claim 1, characterized in that: The distribution network post-disaster maintenance model aims to minimize the system's cumulative load supply shortage and the latest completion time of all maintenance tasks, and comprehensively optimizes the movement path of maintenance personnel, the order of commissioning of repaired transmission lines, and the optimal power flow problem of the power grid.

6. A method for post-disaster maintenance of a distribution network taking into account the dependence of repair time series as claimed in claim 1, characterized in that: The distribution network post-disaster maintenance model constructed by the solution is divided into two stages: the first stage searches for a better candidate maintenance sequence through a heterogeneous dual-population ant colony algorithm based on a collaborative filtering strategy; the second stage optimizes the commissioning order of the transmission lines through a cyclic timing right shift strategy, improves the objective function corresponding to the maintenance sequence, and finally determines the optimal solution to obtain a distribution network post-disaster maintenance plan.

7. A distribution network post-disaster maintenance system taking into account the dependence of repair time series, characterized in that: include: An acquisition module, which is configured to acquire fault point information, traffic network information, and base station information of the distribution network after a natural disaster, and calculate the generalized travel time of each road in the traffic network after the disaster; A determination module, which is configured to determine the shortest repair path considering the post-disaster road damage according to the obtained generalized travel time; A calculation module is configured to calculate the maintenance lead time considering the post-disaster road damage according to the determined shortest maintenance path, and at the same time, calculate the maintenance duration considering the fatigue of maintenance personnel and the real-time impact of extreme weather; A modeling module is configured to combine the obtained maintenance lead time and maintenance duration, take the minimum total cost as the goal, consider the maintenance sequence constraints, component commissioning constraints and distribution network operation constraints, and construct a distribution network post-disaster maintenance model that takes into account the dependence of repair time; The maintenance module is configured to solve the constructed distribution network post-disaster maintenance model, obtain the distribution network post-disaster maintenance plan, and complete the distribution network post-disaster maintenance taking into account the dependence of the repair time series.

8. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the program is executed by a processor, the steps of the method for post-disaster maintenance of a distribution network taking into account the dependency of repair time series as described in any one of claims 1 to 6 are implemented.

9. An electronic device comprising a memory, a processor, and a computer program stored in the memory and running on the processor, characterized in that: When the processor executes the program, the steps of the method for post-disaster maintenance of a power distribution network taking into account the dependency of repair time series according to any one of claims 1 to 6 are implemented.

10. A computer program product comprising software code, characterized in that The program in the software code executes the steps of the method for post-disaster maintenance of a power distribution network taking into account the dependency of a repair time series according to any one of claims 1 to 6.

Images