Emergency material reserve point planning method based on dynamic reconstruction

Abstract

The invention discloses an emergency material reserve point planning method based on dynamic reconstruction, relates to the technical field of post-disaster planning, and solves the problem that the planning of an emergency material reserve point in the prior art does not consider a material in-place sequence and dynamic reconstruction of a power distribution network. The method comprises the following steps: firstly, introducing a deceleration coefficient and a road damage coefficient when calculating road section passing time to reflect the influence of disasters on traffic road capacity; traversing all possible paths, and taking the shortest road section passing time as a screening condition of an optimal path between two nodes; and finally, constructing an objective function based on the construction cost and the maintenance cost of the material storage point and the disaster loss of all typical scenes as influence factors, and solving an optimal planning scheme. According to the invention, material reserve point planning can be carried out by considering different electric power materials and space requirements thereof, the recovery speed of a post-disaster power distribution system can be improved by applying the method to carry out electric power emergency material reserve point planning, and the disaster economic loss is effectively reduced.

Description

Technical Field

[0001] The present invention relates to the technical field of post-disaster planning, and in particular to a method for planning emergency material storage points based on dynamic reconstruction. Background Art

[0002] Against the backdrop of global warming, disastrous weather events are becoming increasingly frequent, causing severe economic losses across all industries. In the power industry, distribution networks, due to their direct access to users and widespread geographical distribution, have long been a primary battleground for disaster relief. With the increasing penetration of renewable energy, the distribution system's vulnerability has further expanded, often facing the risk of tower collapse, line breakage, renewable energy unit failure, and even widespread power outages. To effectively mitigate the impact of extreme weather and enhance the emergency response capabilities of the distribution system, the rational planning of emergency material storage sites has become a critical task. The National Disaster Prevention, Mitigation, and Relief Committee also issued the "Guiding Opinions on Further Strengthening the Emergency Rescue and Disaster Relief Material Guarantee System and Capacity Building," aiming to achieve 100% coverage of county-level storage depots (points) nationwide by the end of 2025.

[0003] Existing research on emergency material storage point planning typically uses the combined minimization of construction cost and load loss as an evaluation metric or objective function, and constructs optimization models that consider uncertainties such as tower and line failures and the extent of road network damage. From a modeling perspective, these approaches can be categorized into three types: 1. Scenario-based stochastic planning uses sampling simulations to generate a large number of typical scenarios to characterize uncertainty. Scenarios are assigned weights to guide the selection of emergency material storage point locations and capacities. Its effectiveness depends on the number of typical scenarios considered. 2. Robust optimization models various uncertainties as uncertainty sets, such as intervals or polyhedrons, and dynamically identifies the worst-case scenario. Reserve point locations and capacities are determined based on this worst-case scenario. 3. Distributionally robust optimization models the possible distributions of uncertain parameters as fuzzy sets, and reserves are planned based on this worst-case distribution. 4. Its effectiveness depends on the construction of the fuzzy sets. From a problem analysis perspective, current research contains several unreasonable assumptions. For example, it assumes that material supply can fully meet the needs of the affected site, that material demand and supply points are unique, and that the transportation network's capacity remains constant. However, reality is often more complex: diverse resource demands may exist, involving multiple supply and demand points, and the transportation network's capacity may also change due to the disaster. Second, existing research has largely focused on the needs of disaster victims, whereas power grid failures present significantly different resource requirements. For example, maintenance personnel must ensure that all emergency supplies are in place before they can begin repairing the fault point. Finally, during the repair process, the power grid may dynamically reconfigure, leading to topological changes and power flow redistribution. These characteristics place higher demands on the planning of emergency reserve points for power supplies and should be key considerations in planning. As the above review indicates, current research on emergency reserve point planning focuses on conventional supplies, and the demanders are typically disaster victims. However, existing methods are not applicable to the planning of emergency reserve points for power supplies for two main reasons: First, a distribution network fault point must be repaired only after all emergency supplies are in place, which differs significantly from the phased nature of disaster victims' resource needs. Second, during the repair process, the distribution network dynamically reconfigures, resulting in changes in power flow distribution, making traditional static topological models inapplicable.

[0004] In view of this, a dynamic reconstruction-based emergency material storage point planning method is needed. Summary of the Invention

[0005] In view of the problem that the planning of emergency material storage points in the existing technology does not take into account the order in which materials arrive and the dynamic reconstruction of the distribution network, the present invention provides an emergency material storage point planning method based on dynamic reconstruction, which can consider different power materials and their spatial requirements for material storage point planning, and through the dynamic recovery process after the disaster of multiple typical scenario models, the storage point planning results are tested and the optimal planning scheme is finally determined. Applying this scheme to plan power emergency material storage points can improve the recovery speed of the power distribution system after the disaster and effectively reduce the economic losses caused by the disaster. The specific technical scheme is as follows:

[0006] A method for planning emergency material storage points based on dynamic reconstruction includes the following steps:

[0007] When calculating the travel time of a road section, a deceleration coefficient and a road damage coefficient are introduced to reflect the impact of disasters on the transportation capacity of traffic roads;

[0008] Traverse all possible paths and use the shortest travel time as the screening condition for the optimal path between two nodes;

[0009] The objective function is constructed based on the construction cost and maintenance cost of the material storage point and the disaster losses in all typical scenarios as influencing factors. The details are as follows:

[0010]

[0011] Where C St represents the construction and maintenance cost of the emergency material storage point; s is the typical fault scenario index, S is the typical fault scenario set; C Loss,s represents the economic loss of the system in the post-disaster stage under scenario s; ω s is the weight of the sth scene;

[0012] Solve the objective function and obtain the optimal planning solution.

[0013] Preferably, the specific calculation method of the road section travel time is as follows:

[0014]

[0015] Where, τ mn is the travel time between intersections m and n; x mn is the traffic volume of the road section; is the zero-flow transit time, k ro is the deceleration coefficient, and k ro ∈(0,1];c mn is the design flow of the road section; α and β are model parameters; k br is the road damage coefficient, and k br ∈(0,1].

[0016] Preferably, the process of obtaining the optimal path is as follows:

[0017] Traverse and record all feasible paths, then calculate and update the road travel time, calculate the travel time of each path, and finally select the shortest path under the current road conditions.

[0018] Optimally, the construction and maintenance cost of the power emergency material storage point C St The specific calculation is as follows:

[0019]

[0020] Where r represents the discount rate; Y represents the land use period; i represents the node index of the power distribution system, and I represents the node set; k represents the type index of power emergency supplies, and K represents the material category set; b is the investment cost of building a material storage point per unit area at node i; k 、 are the floor space and maintenance cost of the k-th material per unit respectively; is the reserve amount of the kth type of emergency supplies in node i.

[0021] Preferably, the economic loss in the post-disaster period C Loss The specific calculation is as follows:

[0022]

[0023] L Br is the set of fault lines; l1, l2, ..., l n Respectively represent the indexes of the first, second, ..., and nth repaired lines; represents the time required to repair line l1, It represents the system's disaster loss per unit time before line l1 is repaired; It represents the time from the repair of line l1 to the repair of the second fault line l2. It represents the system loss per unit time during the period from the repair of line l1 to the repair of the second fault line l2; It represents the time after the first n-1 lines are repaired and before the nth faulty line is repaired. It represents the system loss per unit time after the first n-1 lines are repaired and before the nth faulty line is repaired.

[0024] Preferably, the optimal reconstruction model Ψ is used to determine The value of z is as follows: i,i′ The value of simulates the order in which the lines are repaired, and the objective function value obtained by solving the model Ψ is the corresponding K Loss ;

[0025] The optimal reconstruction model Ψ is as follows:

[0026]

[0027] Where i is the distribution network node index, I is the distribution network node set; is the unit load loss coefficient of node i, P i Load is the active power demand of node i, P i LS is the load loss at node i, is the reactive power demand of node i, is the reactive power reduction at node i, is the importance weight of the node load; (i, i′) is the distribution network line with node i as the starting point and node i′ as the end point, L Sw represents the set of lines with tie switches; K TR It is the risk cost coefficient of the interconnection line transfer; is the active and reactive power flow of line (i, i′); is a binary variable representing the state of the tie switch on line (i, i′); M is a maximum value, ε is a minimum value; P i Grid 、 Indicates the active and reactive power transmitted from the upper transmission network to the distribution network; P i REac 、 is the actual active and reactive output of distributed renewable energy at node i; are respectively the upper limits of active and reactive power transmission of line (i, i′); V i is the voltage at node i, V i′ is the voltage of node i'; V0 is the reference voltage; r i,i′ 、x i,i′ are the resistance and reactance of line (i,i′); V i,min 、V i,max are the minimum and maximum voltages allowed at node i; is the maximum available active power output of the distributed renewable energy at node i; are the maximum and minimum reactive output of distributed renewable energy respectively; N Br is the number of damaged transmission lines.

[0028] Preferably, constraints are also provided, and the constraints include at least material storage point planning constraints, which are as follows:

[0029]

[0030] Where, is a binary variable indicating whether node i deploys an emergency material storage point; is the maximum quantity of the kth type of material that a single storage point is allowed to store; It is the maximum number of reserve points allowed to be built.

[0031] Preferably, the solution algorithm for obtaining the optimal planning solution is as follows:

[0032] Step 1: Parameter initialization: Input the distribution system topology, node active and reactive power requirements, node load weights, installed capacity of new energy generators, transportation network topology, design parameters of each road section, and historical average traffic statistics;

[0033] Step 2: Generate typical fault scenarios: First, analyze historical data from each disaster to determine the number of faults on each line and tower, converting this number into a probability of failure. Similarly, the extent of damage to the road is determined. Next, use Monte Carlo simulation to generate a large number of fault scenarios. Finally, use K-means clustering to select S typical fault scenarios, and assign scenario weights based on the number of samples in each cluster.

[0034] Step 3: Initialize PSO algorithm parameters and particle information: The information carried by the particles is the location of the emergency material storage point and the quantity of various types of material reserves. The dimension of the particle is

[0035] Step 4: Calculate the differentiated parameters under different scenarios: For each typical fault scenario, simulate different line repair orders, solve the model Ψ, and obtain Furthermore, based on the traffic road and flow information in the current scenario, the shortest travel time between nodes is calculated according to the method proposed in the first part;

[0036] Step 5: Calculation of disaster losses in each scenario: For each typical fault scenario, solve the power emergency material storage point planning model with equation (4) as the objective function and equations (7)-(15) as constraints;

[0037] Step 6: Update particle information: Calculate formula (3) based on the reserve point information carried by the particle itself, and combine it with the result calculated in step 5 to calculate formula (2) as the fitness; update the individual optimal value, global optimal value, particle speed and particle position;

[0038] Step 7: Determine whether the iteration is finished: Determine whether the set number of iterations has been reached. If so, output the particle information corresponding to the current global optimal value as the result of the emergency material storage point planning; otherwise, return to step (4).

[0039] A computer-readable storage medium includes a stored program, wherein when the program is run, the device where the computer-readable storage medium is located is controlled to execute the emergency material storage point planning method based on dynamic reconstruction as described above.

[0040] A processor is used to run a program, wherein when the program is run, the emergency material storage point planning method based on dynamic reconstruction as described above is executed.

[0041] Compared with the prior art, the present invention has the following beneficial effects:

[0042] The present invention first introduces a deceleration coefficient and a road damage coefficient when calculating the travel time of a road section to reflect the impact of disasters on traffic road capacity; then, all possible paths are traversed, with the shortest road section travel time as the screening condition for the optimal path between two nodes; finally, with the comprehensive optimization of the construction cost of emergency material storage points and post-disaster system losses as the goal, the objective function is constructed based on the construction cost of the material storage points, the maintenance cost, and the disaster losses of all typical scenarios as influencing factors, and the optimal planning scheme is solved. The present invention performs data preprocessing, model construction, and model solving, and their sequential execution can achieve reasonable planning of power material storage points, effectively improve the post-disaster recovery capability of the distribution system, and reduce the economic losses caused by load loss. BRIEF DESCRIPTION OF THE DRAWINGS

[0043] To more clearly illustrate the specific embodiments of the present invention or the technical solutions in the prior art, the following briefly describes the drawings required for the specific embodiments or the description of the prior art. Similar elements or parts are generally identified by similar reference numerals throughout the drawings. Elements or parts in the drawings are not necessarily drawn to scale.

[0044] Figure 1 This is a schematic diagram of the depth-first algorithm;

[0045] Figure 2 Schematic diagram of disaster loss assessment taking into account dynamic network reconstruction;

[0046] Figure 3 Flowchart of the algorithm for solving the two-tier MES pre-layout model based on PSO;

[0047] Figure 4 To test the topology of the 35-node power distribution system;

[0048] Figure 5 The traffic network topology corresponding to the test 35-node power distribution system;

[0049] Figure 6 Schematic diagram of the average annual disaster losses and average restoration time for each case. DETAILED DESCRIPTION

[0050] The following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of them. All other embodiments obtained by ordinary technicians in this field based on the embodiments of the present invention without making any creative efforts shall fall within the scope of protection of the present invention.

[0051] It will be understood that when used in this specification and the appended claims, the terms “comprises” and “comprising” indicate the presence of described features, integers, steps, operations, elements and / or components, but do not preclude the presence or addition of one or more other features, integers, steps, operations, elements, components and / or groups thereof.

[0052] It should also be understood that the terms used in the present specification are only for the purpose of describing particular embodiments and are not intended to limit the present invention. As used in the present specification and the appended claims, the singular forms "a", "an", and "the" are intended to include the plural forms unless the context clearly indicates otherwise.

[0053] It should be further understood that the term "and / or" used in the present description and the appended claims refers to and includes any and all possible combinations of one or more of the associated listed items.

[0054] The technical approach of this solution consists of three parts. The first part constructs a method for identifying optimal movement paths for maintenance teams based on the Bureau of Public Roads Model (BPR) and a depth-first algorithm. The second part proposes a planning model for power emergency material storage points that takes into account grid-transportation network coupling and dynamic reconfiguration. The third part describes the solution algorithm for this planning model. The first part considers the impact of disasters on road capacity, calculates road travel times based on the BPR model, and uses a depth-first algorithm to identify all feasible paths between two nodes, calculates the travel time for each path, and ultimately selects the optimal movement path, providing input parameters for the subsequent emergency material storage point planning model that takes into account grid-transportation network coupling. The second part considers fault uncertainty and distribution system network reconfiguration during the recovery process, and establishes a stochastic programming (SP)-based planning model for power emergency material storage points, aiming to optimize the combined cost of emergency material storage point construction and post-disaster system losses. The third part uses the input parameters from the first part to iteratively solve the emergency material storage point planning model using the particle swarm optimization (PSO) algorithm to obtain the optimal location and sizing solution. These three parts involve data preprocessing, model construction, and model solution. Their sequential execution enables the rational planning of power material storage points, effectively improving the post-disaster recovery capabilities of the distribution system and reducing economic losses caused by load loss.

[0055] The innovations of this solution are: 1) taking into full account the repair logic of power failures, a method for planning power emergency material storage points suitable for power distribution systems affected by disasters is proposed; 2) a power emergency material distribution model is constructed that takes into account the optimal network reconstruction during post-disaster recovery, which can provide a more reasonable material distribution plan.

[0056] 1. Optimal Movement Path Identification Method for Maintenance Teams Based on BPR Model and Depth-First Algorithm

[0057] 1. Road travel time calculation model based on BPR model

[0058] BPR, an empirical model proposed by the U.S. Bureau of Public Roads, describes the nonlinear relationship between traffic flow and travel time. This proposal further considers the impact of disasters on road capacity based on the original model and constructs the following improved BPR model:

[0059]

[0060] Where, τ mn is the travel time between intersections m and n; x mn is the traffic volume of the road section; is the zero-flow travel time, which is related to the length of the road section and the design speed. However, after a disaster, drivers usually drive at a slower speed than usual for safety. Therefore, this scheme considers a deceleration coefficient k ro , k ro ∈(0,1].c mn is the design flow of the road section; α and β are model parameters, usually taking values ​​of 0.15 and 4; k br is the road damage coefficient, k br ∈(0,1], which depends on the construction of the road. For example, if a roadside tree falls, the narrower the road, the greater the impact, or even completely impassable, that is, k br →0.

[0061] 2. Shortest path identification based on depth-first algorithm

[0062] The choice of the shortest path identification method is closely related to the research object. The urban power distribution system has a high degree of cabling and a low penetration rate of new energy, and is relatively weakly affected by disasters. Therefore, this scheme chooses the rural new energy distribution system with a wider distribution area, higher new energy penetration rate, and more serious disaster impact as the research object. Since rural roads are usually undirected, breadth-first search (BFS), depth-first search (DFS), and Dijkstra algorithm can all be used for shortest path identification. When the road state remains unchanged, breadth-first search and Dijkstra algorithm can effectively identify the shortest path. However, disasters can cause changes in road travel time, thereby affecting the shortest path between two nodes. At this time, if the algorithm is repeatedly called to search for the shortest path, the amount of calculation may increase. To address this problem, this scheme considers using the DFS algorithm. The basic idea of ​​​​DFS is to continue to go deeper along a branch until it can no longer continue, then backtrack to the previous node and try other branches until all possible paths are explored, such as Figure 1 shown.

[0063] DFS can traverse and record all feasible paths, then update the road travel time based on formula (1), calculate the travel time of each path, and finally select the shortest path under the current road conditions to guide the travel of the maintenance team. The advantage of this method is that it only needs to call DFS once, effectively reducing repeated calculations. The travel time calculated in this section (hereinafter represented as t OtoD ) is the key parameter of the subsequent material allocation model.

[0064] 2. Planning Model for Power Emergency Material Reserve Points Considering Grid-Transportation Network Coupling and Dynamic Reconfiguration

[0065] 1. Objective Function

[0066] The main goal of planning power emergency material storage points is to reduce economic losses after disasters. Since failures caused by disasters are uncertain information, this solution adopts scenario-based stochastic planning to build a model. Typical scenarios are obtained by first obtaining the number of failures of each line and tower through historical data, converting it into the probability of failure, then using Monte Carlo simulation to generate a large number of failure scenarios, and finally using K-means clustering to screen out S typical failure scenarios. The goal of this solution is to minimize the construction cost, maintenance cost, and disaster losses of material storage points for all typical scenarios while considering the weight of each typical scenario, as shown below:

[0067]

[0068] Where C St represents the construction and maintenance cost of the emergency material storage point; s is the typical fault scenario index, S is the typical fault scenario set; C Loss,s represents the economic loss of the system in the post-disaster period under scenario s. The post-disaster period starts from the end of the disaster and the time when the maintenance team can be dispatched, and ends when the system is restored; ω s is the weight of the sth scene. C St and C in a single scenario Loss is calculated as follows:

[0069] 1) Construction and maintenance costs of power emergency material storage points C St

[0070]

[0071] Where r represents the discount rate, which is 5% here; Y represents the land use period, which is 50 years here; i represents the node index of the distribution system, and I is the node set; k represents the type index of power emergency supplies, and K is the material category set; b is the investment cost of building a material storage point per unit area at node i, taking into account land price and construction material costs; k 、 are the floor space and maintenance cost of the k-th material per unit respectively; is the reserve amount of the kth type of emergency supplies in node i.

[0072] 2) Economic losses after disasters C Loss

[0073] This solution takes into account dynamic network reconstruction and conducts phased disaster loss assessment. After each line is repaired, the system will perform an optimal network reconstruction to ensure minimal load loss until the network is restored. Figure 2 The loss assessment process of this scheme is demonstrated by taking the failure of three lines as an example.

[0074] correspond Figure 2 The objective function of the disaster loss assessment process is as follows:

[0075]

[0076] Where, L Br is the set of faulty lines; l1, l2, and l3 represent the indexes of the first, second, and third repaired lines, respectively. It represents the time required to repair line l1. This solution briefly describes it as the first stage. It represents the system's disaster loss per unit time before line l1 is repaired; is the time it takes for all materials needed for line l1 maintenance to arrive. These are the time taken to repair lines L1, L2, and L3 after the materials arrive. It represents the time from the repair of line l1 to the repair of the second fault line l2, which is briefly described as the second stage. It represents the system's disaster loss per unit time during this period. It indicates the time it takes for all materials on line L2 to arrive after line L1 is repaired. It represents the time from the time when the first and second fault lines are repaired to the time when the third fault line is repaired, which is briefly referred to as the third stage. It represents the system's disaster loss per unit time during this period. This represents the time it takes from the time line L2 is repaired to the time all supplies are in place at L3. If there are more faulty lines, the above process can be further extended.

[0077] These three parameters are crucial to determining the order of material distribution. This solution uses the optimal reconstruction model Ψ to determine their values, as shown below:

[0078]

[0079] In the objective function of model Ψ, i is the distribution network node index, and I is the set of distribution network nodes; is the unit load loss coefficient of node i, P i Load is the active power demand of node i, P i LS is the load loss at node i, is the reactive power demand of node i, is the reactive power reduction at node i, is the importance weight of the node load; (i, i′) is the distribution network line with node i as the starting point and node i′ as the end point, LSw represents the set of lines with tie switches; K TR This is the risk cost coefficient for tie-line power transfer. The greater the power transferred by the tie-line, the more likely it is to cause overload, thus posing a potential risk. is the active and reactive power flow of line (i, i′); is a binary variable representing the state of the tie switch on line (i, i′) (0 means the switch is open, 1 means the switch is closed); M is a maximum value, and this paper takes 1e 4 ,ε is a minimum value, and this paper takes 1e -4 The objective function (5.1) is to minimize the load loss based on the reconfigured distribution network. The second term is used to select the option that makes the load of all interconnection lines as low as possible when there are multiple feasible reconstruction options, so as to avoid line overload and reduce the pressure of power transfer. The third term is used to reduce unnecessary interconnection switch actions to avoid closed-loop operation of the system.

[0080] Equations (5.2)-(5.3) are the power balance constraints of the distribution system. This solution uses linearized distflow to characterize them, where: P i Grid 、 Indicates the active and reactive power transmitted from the upper transmission network to the distribution network; P i REac 、 is the actual active and reactive output of distributed renewable energy at node i.

[0081] Equations (5.4) to (5.8) represent the line state and line transmission power constraints, where: are the upper limits of active and reactive power transmission of line (i, i′); L represents the set of all transmission lines; z i,i′ is a binary variable indicating whether the lines (i, i′) are connected (0 for disconnected, 1 for connected). Except for the line where the tie switch is located, the z values ​​of the other lines are i,i′ It can be pre-set according to the current line fault situation, and the z of the tie switch i,i′ By optimizing Make confirmation.

[0082] Equations (5.9)-(5.10) are voltage constraints, where: V i is the voltage at node i, V i′ is the voltage of node i'; V0 is the reference voltage; r i,i′ 、x i,i′ are the resistance and reactance of line (i,i′); V i,min 、V i,max are the minimum and maximum voltages allowed at node i.

[0083] Equations (5.11) and (5.12) are the power constraints of distributed renewable energy, where: is the maximum available active power output of the distributed renewable energy at node i; are the maximum and minimum reactive output of distributed new energy, respectively. It can be less than 0 and is used to absorb reactive power to regulate voltage.

[0084] Formula (5.13) is the load loss constraint. Formula (5.14) is the radial network constraint, N Br is the number of damaged transmission lines. This constraint is also used to maintain open-loop operation after the distribution network is reconfigured.

[0085] This scheme adjusts z i,i′ The value of simulates the order in which the lines are repaired, and the objective function value obtained by solving the model Ψ is the corresponding K Loss It should be noted that in model Ψ, this scheme uses (i, i′) to represent the line to facilitate the calculation of node power balance and line power flow, while in equation (4), this scheme uses l to represent the line to reduce the variable dimension. These two representation methods do not conflict, and we can number the lines to achieve a one-to-one correspondence between (i, i′) and l.

[0086] 2. Constraints

[0087] 1) Material storage point planning constraints

[0088]

[0089] Where, is a binary variable indicating whether node i deploys an emergency material storage point; is the maximum quantity of the kth type of material that a single storage point is allowed to store; It is the maximum number of reserve points allowed to be built.

[0090] The following constraints are used to optimize resource allocation under a certain failure scenario and then calculate the post-disaster economic losses.

[0091] 2) Material demand constraints

[0092]

[0093] Where l is the line index; y i,l,k is the quantity of the kth material transported by node i to line l; are the gap and demand of the kth material on fault line l respectively.

[0094] 3) Line status judgment constraints

[0095]

[0096] Where, Is a binary variable indicating whether the fault line has obtained all the required supplies from the emergency material reserve point. If the line l does not get enough supplies due to the limited reserves of the material reserve point, then Otherwise it is 0. Br} means that set L excludes set L Br The set consisting of the remaining elements after the contained elements, M1 is a maximum value, and "\" is the symbol used in mathematics to exclude set elements.

[0097] 4) Balance constraints between material storage and delivery

[0098]

[0099] Where, is the reserve amount of the kth material at node i.

[0100] 5) Transportation time constraints

[0101]

[0102] Where, Indicates the time taken for each fault line to be repaired; It is a binary variable indicating whether node i delivers materials to line l, 1 means delivery, 0 means no delivery; represents the time required for node i to deliver materials to line l, which can be obtained from the first part; T OC Indicates the time required to transfer supplies from other cities when the local storage point is insufficient; T l Rp Represents the time required to repair the faulty line. Note that in this solution, l, l1, l2, and l3 are all line indices, but with numerical subscripts added to distinguish them at different stages. l1′, l2′, and l3′ represent line indices that are different from l1, l2, and l3, respectively, and are used to construct related constraints. Note that in this solution, l, l1, l2, and l3 are all line indices, but with numerical subscripts added to distinguish them at different stages.

[0103] 6) Relevant constraints in the first stage

[0104]

[0105] Where, It represents the time required to repair line l1. is a binary variable used to index the first faulty line repaired, Indicates that the line l1 is repaired first; The first constraint of formula (13) indicates that there is only one line to be repaired first, and the second constraint is used to determine The third constraint is used to determine the value of The value of , the fourth constraint determines The value of M2 is used to calculate the duration of the subsequent stages. M2 is a maximum value smaller than M1.

[0106] 7) Second stage related constraints

[0107]

[0108] Where, is a binary variable used to index the second repaired fault line. The calculation formula is the repair time of each line minus the duration of the first stage. In other words, Indicates the remaining repair time of the remaining faulty lines after the first line is repaired. Here, in order to avoid the repaired lines interfering with the judgment of the repair order of subsequent lines, this solution adopts This item sets the remaining repair time of the repaired line to a maximum value. Indicates the minimum remaining repair time of all faulty lines in the second phase, excluding the repaired lines. Its value is equal to the duration of the second phase.

[0109] 8) Relevant constraints of the third stage

[0110]

[0111] Where, is a binary variable used to index the third (last) repaired fault line; It represents the remaining repair time for the third line after the first and second lines are repaired. It should be noted that if there are more than three faulty lines, only the second stage constraint form needs to be used for stage expansion.

[0112] At this point, the planning model for power emergency material storage points taking into account the coupling of distribution network and transportation network and dynamic network reconstruction has been completed.

[0113] 3. PSO-based electric power emergency material storage point planning solution algorithm

[0114] The model proposed in the second section is a mixed integer programming model. When the system under consideration is small and the number of typical scenarios is low, this model can be solved directly using commercial solvers (such as CPLEX and Gorubi). However, for reasonable and scientific planning, it is essential to analyze as many typical scenarios as possible. This results in an exponential increase in the problem size, which in turn increases memory requirements and significantly prolongs the solution time.

[0115] In order to reduce the demand for memory, this solution proposes a PSO-based emergency material storage point planning and solving algorithm. The specific process is as follows:

[0116] Step 1: Parameter initialization. Input data such as the distribution system topology, node active and reactive power requirements, node load weights, installed capacity of new energy generators, transportation network topology, design parameters of each road section, and historical average traffic statistics.

[0117] Step 2: Generate typical failure scenarios. First, analyze historical data from each disaster to determine the number of failures for each line and tower, converting this number into a probability of failure. Similarly, analyze the extent of damage to the road. Then, use Monte Carlo simulation to generate a large number of failure scenarios. Finally, use K-means clustering to select S typical failure scenarios, and assign scenario weights based on the number of samples in the cluster.

[0118] Step 3: Initialize PSO algorithm parameters and particle information. The information carried by the particles is the location of emergency material storage points and the quantity of various types of material reserves. The dimension of the particles is

[0119] Step 4: Calculate the differential parameters under different scenarios. For each typical fault scenario, simulate different line repair orders and solve the model Ψ to obtain In addition, based on the traffic road and flow information in this scenario, the shortest travel time between nodes is calculated according to the method proposed in the first part.

[0120] Step 5: Calculation of disaster losses in each scenario. For each typical fault scenario, solve the power emergency material storage point planning model with equation (4) as the objective function and equations (7)-(15) as constraints.

[0121] Step 6: Update particle information. Calculate equation (3) based on the reserve point information carried by the particle itself, and combine it with the result calculated in step 5 to calculate equation (2) as the fitness; update the individual optimal value, global optimal value, particle speed, and particle position.

[0122] Step 7: Determine whether the iteration is complete. Determine whether the set number of iterations has been reached. If so, output the particle information corresponding to the current global optimal value as the result of the emergency material storage point planning; otherwise, return to step (4).

[0123] The diagram of the solution process is as follows Figure 3 shown.

[0124] In order to verify the feasibility and effectiveness of this scheme, the 35-node distribution network system of Babai Line in Fangchenggang City, Guangxi Province was used for case testing.

[0125] 1. Test system data description

[0126] Figure 4 The topology of the Babai Line distribution network in Fangchenggang City, Guangxi Province is shown. Node 4 is equipped with a 5MW centralized photovoltaic power station, node 7 is equipped with a 1.5MW wind turbine, and nodes 13, 21, and 33 are equipped with rooftop photovoltaics with capacities of 1.2MW, 1.5MW, and 1.2MW, respectively. Nodes marked with black triangles in the figure indicate that the loads on them are relatively important. The remaining nodes The distribution system consists of three interconnecting lines: TS1, TS2, and TS3. During the test, the load loss was set at 6.5 yuan / kWh, and the interconnecting line transfer penalty coefficient was set at 0.5 yuan / kWh.

[0127] Figure 5 The traffic network topology diagram corresponding to the 35-node distribution system tested is shown. In the diagram, the node numbers correspond to the node numbers on the distribution network. The numbers on the lines represent the road length, the line width indicates the road grade, and dashed lines indicate roads that are impassable for large vehicles. Note that due to the winding nature of some roads, this solution straightened them when drawing them, resulting in a discrepancy between the lengths of the lines on the diagram and the actual road lengths.

[0128] Based on 10 years of equipment disaster failure data in the region, this plan calculated that there are an average of 8 severe disaster weather events per year. It also counted the probability of transmission channels between nodes being damaged and the demand for emergency power supplies. Subsequently, a large number of failure scenarios were generated using Monte Carlo simulation, and 10 typical failure scenarios were selected through K-means clustering. This plan assumes that the distribution network will invest in a maximum of two emergency material storage points, with a design height of 6 meters and a land acquisition and construction cost of 8,000 yuan per square meter. 2 After the discount rate is converted, the equivalent annual construction cost is 438.4 yuan / m 2 Emergency supplies are divided into three categories. The first category is large supplies (such as transformers, towers, etc.), which occupy an average of about 3m 2 , maintenance costs 200 yuan per piece of equipment per year; the second category of materials is medium-sized materials (such as circuit breakers, disconnectors, etc.), which occupy an average of about 2m 2 , maintenance costs 100 yuan per piece of equipment per year; the third category of materials is small materials (such as wires, insulators, etc.), which occupy an average of about 0.2m 2 , maintenance costs The cost is 5 yuan per unit per year. Furthermore, if local emergency power supplies are insufficient to support repairs, assistance will need to be requested from neighboring regions. Considering the limited availability of personnel and transportation capacity after a disaster, this plan assumes that support supplies will arrive within approximately four hours.

[0129] This solution uses three different cases to verify the effectiveness of the solution, as shown below:

[0130] Case 1: This plan is used to select the site and capacity of power emergency material storage points.

[0131] Case 2: No storage site selection is carried out, and the materials are placed in the upper-level 35kV substation, and only the material storage volume is planned.

[0132] Case 3: Site selection and capacity determination of power emergency material storage points without considering dynamic reconstruction.

[0133] Table 1 shows the planning results for emergency material storage points in each case. Based on the planning results in Table 1, this solution used Monte Carlo simulation to generate 4,000 fault scenarios, grouping them into groups of 400 to simulate the impact of a disaster on the distribution network over a 50-year period. A total of 10 test groups were designed, and the average of the test results was taken to reduce sampling contingency. Figure 6 The disaster losses and average recovery time for the three cases are shown.

[0134] Table 1 Planning schemes for emergency material storage points in each case

[0135]

[0136] Comparing the results of Case 1 and Case 2, we can see that because Case 2 lacked planning for the location of emergency material storage points, maintenance teams had to travel farther to reach the fault point, increasing the system's average annual disaster losses and average recovery time. Therefore, it is essential to consider the optimal location of emergency material storage points.

[0137] Comparing Case 1 and Case 3 reveals that Case 3's decision to reserve more emergency power supplies resulted in an increase of 13,500 yuan (43.2%) in annualized investment and maintenance costs compared to Case 1. This increase was primarily due to Case 3's failure to consider dynamic reconfiguration during the maintenance process and its inability to fully utilize interconnection lines for power transfer. Consequently, the system overestimated disaster losses, leading to a conservative planning of emergency supply storage locations. Although higher emergency supply reserves reduced the need for support from neighboring regions and significantly improved the average time to recovery, the reduction in disaster losses was only 3,600 yuan. Combined with the increase in investment and maintenance costs, the cost-effectiveness of this additional investment is low. Furthermore, some faults can be avoided through power transfer, requiring less immediate repairs. Even if these cannot be quickly restored, they do not result in significant economic losses. Therefore, considering dynamic reconfiguration during maintenance is crucial for accurately assessing system losses and selecting the most cost-effective emergency supply storage location plan.

[0138] In summary, the present invention consists of three parts. The first part constructs a method for identifying the optimal mobile path of maintenance teams based on the Bureau of Public Roads Model (BPR) model and the depth-first algorithm. The second part proposes a planning model for power emergency material storage points that takes into account the coupling and dynamic reconstruction of the power grid and transportation network. The third part describes the solution algorithm of the planning model. The function of the first part is to consider the impact of disasters on road capacity, calculate the road travel time based on the BPR model, and identify all feasible paths between two nodes through the depth-first algorithm, calculate the travel time of each path, and finally select the optimal mobile path to provide input parameters for the subsequent emergency material storage point planning model that takes into account the coupling of the power grid and transportation network. The function of the second part is to consider the uncertainty of faults and the reconstruction of the distribution system network during the recovery process, and establish a power emergency material storage point planning model based on stochastic programming (SP) with the goal of comprehensively optimizing the construction cost of emergency material storage points and post-disaster system losses. The third part uses the input parameters from the first part to iteratively solve the emergency material storage point planning model using the particle swarm optimization (PSO) algorithm to obtain the optimal location and sizing solution. These three parts involve data preprocessing, model construction, and model solution. Their sequential execution enables the rational planning of power material storage points, effectively improving the post-disaster recovery capabilities of the distribution system and reducing economic losses caused by load loss.

[0139] Those skilled in the art will appreciate that the units of the various examples described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, computer software, or a combination of the two. In order to clearly illustrate the interchangeability of hardware and software, the composition of each example has been generally described in terms of function in the above description. Whether these functions are performed in hardware or software depends on the specific application and design constraints of the technical solution. Professional and technical personnel can use different methods to implement the described functions for each specific application, but such implementation should not be considered to be beyond the scope of the present invention.

[0140] In the embodiments provided by the present invention, it should be understood that the division of units is merely a logical function division, and there may be other division methods in actual implementation, for example, multiple units can be combined into one unit, one unit can be split into multiple units, or some features can be ignored, etc.

[0141] In addition, the functional units in the various embodiments of the present invention may be integrated into a single processing unit, each unit may exist physically separately, or two or more units may be integrated into a single unit. The aforementioned integrated units may be implemented in the form of hardware or software functional units.

[0142] If the integrated unit is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, or the part that contributes to the prior art, or all or part of the technical solution can be embodied in the form of a software product. The computer software product is stored in a storage medium and includes several instructions for enabling a computer device (which can be a personal computer, server or network device, etc.) to perform all or part of the steps of the method described in each embodiment of the present invention. The aforementioned storage medium includes: U disk, read-only memory (ROM, Read-0nly Memory), random access memory (RAM, Random Access Memory), mobile hard disk, magnetic disk or optical disk, etc., various media that can store program code.

[0143] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit it. Although the present invention has been described in detail with reference to the above embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the above embodiments, or make equivalent replacements for some or all of the technical features therein. These modifications or replacements do not deviate the essence of the corresponding technical solutions from the scope of the technical solutions of the embodiments of the present invention, and they should all be included in the scope of the claims and description of the present invention.

Claims

1. A method for planning emergency material storage points based on dynamic reconstruction, characterized in that: The following steps are involved: When calculating the travel time of a road section, a deceleration coefficient and a road damage coefficient are introduced to reflect the impact of disasters on the transportation capacity of traffic roads; Traverse all possible paths and use the shortest travel time as the screening condition for the optimal path between two nodes; The objective function is constructed based on the construction cost and maintenance cost of the material storage point and the disaster losses in all typical scenarios as influencing factors. The details are as follows: Where C St represents the construction and maintenance costs of emergency material storage points; s is the typical fault scenario index, S is the typical fault scenario set; C Loss,s represents the economic loss of the system in the post-disaster stage under scenario s; ω s is the weight of the sth scene; Solve the objective function and obtain the optimal planning solution.

2. The method for planning emergency material storage points based on dynamic reconstruction according to claim 1 is characterized in that: The specific calculation method of road section travel time is as follows: Where, τ mn is the travel time between intersections m and n; x mn is the traffic volume of the road section; is the zero-flow transit time, k ro is the deceleration coefficient, and k ro ∈(0,1];c mn is the design flow of the road section; α and β are model parameters; k br is the road damage coefficient, and k br ∈(0,1].

3. The method for planning emergency material storage points based on dynamic reconstruction according to claim 1, characterized in that: The process of obtaining the optimal path is as follows: Traverse and record all feasible paths, then calculate and update the road travel time, calculate the travel time of each path, and finally select the shortest path under the current road conditions.

4. The method for planning emergency material storage points based on dynamic reconstruction according to claim 1, characterized in that: Construction and maintenance costs of power emergency material storage points C St The specific calculation is as follows: Where r represents the discount rate; Y represents the land use period; i represents the node index of the distribution system, and I represents the node set; k represents the type index of power emergency supplies, and K represents the material category set; b is the investment cost of building a material storage point per unit area at node i; k 、 are the floor space and maintenance cost of the k-th material per unit respectively; is the reserve amount of the kth type of emergency supplies in node i.

5. The method for planning emergency material storage points based on dynamic reconstruction according to claim 1 is characterized in that: Economic losses in the post-disaster period Loss The specific calculation is as follows: L Br is the set of fault lines; l1, l2, ..., l n Respectively represent the indexes of the first, second, ..., and nth repaired lines; represents the time required to repair line l1, It represents the system's disaster loss per unit time before line l1 is repaired; It represents the time from the repair of line l1 to the repair of the second fault line l2. It represents the system loss per unit time during the period from the repair of line l1 to the repair of the second fault line l2; It represents the time after the first n-1 lines are repaired and before the nth faulty line is repaired. It represents the system loss per unit time after the first n-1 lines are repaired and before the nth faulty line is repaired.

6. The method for planning emergency material storage points based on dynamic reconstruction according to claim 5 is characterized in that: Determined by the optimal reconstruction model Ψ The value of z is as follows: i,i′ The value of simulates the order in which the lines are repaired, and the objective function value obtained by solving the model Ψ is the corresponding K Loss ; The optimal reconstruction model Ψ is as follows: Where i is the distribution network node index, I is the distribution network node set; is the unit load loss coefficient of node i, P i Load is the active power demand of node i, P i LS is the load loss at node i, is the reactive power demand of node i, is the reactive power reduction at node i, is the importance weight of the node load; (i,i′) is the distribution network line with node i as the starting point and node i′ as the end point, L Sw represents the set of lines with tie switches; K TR It is the risk cost coefficient of the interconnection line transfer; is the active and reactive power flow of line (i, i′); is a binary variable representing the state of the tie switch on line (i, i′); M is a maximum value, ε is a minimum value; P i Grid 、 Indicates the active and reactive power transmitted from the upper transmission network to the distribution network; P i REac 、 is the actual active and reactive output of distributed renewable energy at node i; are respectively the upper limits of active and reactive power transmission of line (i, i′); V i is the voltage at node i, V i′ is the voltage of node i'; V0 is the reference voltage; r i,i′ 、x i,i′ are the resistance and reactance of line (i,i′); V i,min 、V i,max are the minimum and maximum voltages allowed at node i; is the maximum available active power output of the distributed renewable energy at node i; are the maximum and minimum reactive output of distributed renewable energy respectively; N Br is the number of damaged transmission lines.

7. The method for planning emergency material storage points based on dynamic reconstruction according to claim 1, characterized in that: Constraints are also set, which at least include material storage point planning constraints, as follows: Where, is a binary variable indicating whether node i deploys an emergency material storage point; is the maximum quantity of the kth type of material that a single storage point is allowed to store; It is the maximum number of reserve points allowed to be built.

8. The method for planning emergency material storage points based on dynamic reconstruction according to claim 1, characterized in that: The solution algorithm for obtaining the optimal planning solution is as follows: Step 1: Parameter initialization: Input the distribution system topology, node active and reactive power requirements, node load weights, installed capacity of new energy generators, transportation network topology, design parameters of each road section, and historical average traffic statistics; Step 2: Generate typical fault scenarios: First, analyze historical data from each disaster to obtain the number of faults on each line and tower, convert this number into a probability of fault occurrence, and similarly, obtain the degree of damage to the traffic road. Then, a large number of fault scenarios are generated using Monte Carlo simulation. Finally, K-means clustering is used to screen out S typical fault scenarios, and the scenario weights are assigned based on the number of samples in the cluster. Step 3: Initialize PSO algorithm parameters and particle information: The information carried by the particles is the location of the emergency material storage point and the quantity of various types of material reserves. The dimension of the particle is Step 4: Calculate the differentiated parameters under different scenarios: For each typical fault scenario, simulate different line repair orders, solve the model Ψ, and obtain Furthermore, based on the traffic road and flow information in the current scenario, the shortest travel time between nodes is calculated according to the method proposed in the first part; Step 5: Calculation of disaster losses in each scenario: For each typical fault scenario, solve the power emergency material storage point planning model with equation (4) as the objective function and equations (7)-(15) as constraints; Step 6: Update particle information: Calculate formula (3) based on the reserve point information carried by the particle itself, and combine it with the result calculated in step 5 to calculate formula (2) as the fitness; update the individual optimal value, global optimal value, particle speed and particle position; Step 7: Determine whether the iteration is finished: Determine whether the set number of iterations has been reached. If so, output the particle information corresponding to the current global optimal value as the result of the emergency material storage point planning; otherwise, return to step (4).

9. A computer-readable storage medium, characterized in that The computer-readable storage medium includes a stored program, wherein when the program is running, the device where the computer-readable storage medium is located is controlled to execute the emergency material storage point planning method based on dynamic reconstruction according to any one of claims 1 to 8.

10. A processor, characterized in that: The processor is used to run a program, wherein the program, when running, executes the emergency material storage point planning method based on dynamic reconstruction according to any one of claims 1 to 8.

Images