Power grid cable wiring path optimization method and device

Abstract

The invention provides a power grid cable wiring path optimization method and device, and the method comprises the steps: carrying out the load analysis of geographical distribution data and power utilization data of a power grid system, and generating load region types and weight data corresponding to each region type; performing fault propagation risk assessment on power grid nodes in the power grid system according to the load area type and the weight data, and generating a cable capacity scheme according to a risk assessment result; generating a candidate path set through a multi-target balance mechanism and an adaptive path growth algorithm according to the load area type, the weight data and a cable capacity scheme; and carrying out toughness evaluation and optimization on the candidate path set under various preset simulation scene conditions to generate a power grid cable wiring scheme. According to the method, the power grid load distribution characteristic, the fault propagation risk and the system toughness are comprehensively considered, a cable wiring scheme which adapts to a complex load environment and has fault resistance can be generated, and the reliability and stability of a power grid system are remarkably improved.

Description

Technical Field

[0001] The present invention relates to the technical field of cable wiring, and in particular to a method and device for optimizing power grid cable wiring paths. Background Art

[0002] Existing methods for optimizing power grid cable routing primarily employ single-objective optimization strategies, typically focusing solely on minimizing total cost or shortest path distance. These strategies lack comprehensive consideration of the grid's actual load distribution, the risk of fault propagation, and system resilience in extreme situations. Particularly in complex terrain and variable load environments, traditional routing methods struggle to cope with the dynamic temporal and spatial distribution of grid loads. This can easily lead to inappropriate cable capacity allocation and a lack of emergency response capabilities in the face of regional faults, effectively hindering the reliability and stability of the entire power grid system. Summary of the Invention

[0003] The main purpose of this invention is to solve the technical problem of ignoring load distribution characteristics, fault propagation risks and system resilience in existing power grid cable routing optimization methods; A first aspect of the present invention provides a method for optimizing a power grid cable wiring path, the method comprising: Perform load analysis on the geographical distribution data and electricity consumption data of the power grid system to generate load area types and weight data corresponding to each area type; performing a fault propagation risk assessment on grid nodes in the power grid system based on the load area type and weight data, and generating a cable capacity plan based on the risk assessment results; Generate a set of candidate paths based on the load area type and weight data and the cable capacity plan through a multi-objective balancing mechanism and an adaptive path growing algorithm; The candidate path set is evaluated for resilience and optimized under various preset simulation scenarios to generate a power grid cable wiring plan.

[0004] Optionally, in a first implementation of the first aspect of the present invention, performing load analysis on the geographical distribution data and electricity consumption data of the power grid system to generate load area types and weight data corresponding to each area type includes: Collecting geographical distribution data and electricity consumption data in the power grid system, wherein the geographical distribution data includes the location of substations, the location of distribution stations, and the node location of each electricity consumption node; Dividing the power consumption data into time periods, calculating the load value of each power consumption node in different time periods, and constructing a spatiotemporal distribution matrix of the power consumption node load; Calculating the load center of gravity position in different time periods based on the spatiotemporal distribution matrix of the power node load and the geographical distribution data, and analyzing the drift trajectory of the load center of gravity position; According to the drift trajectory, the power grid area is divided into three load area types: a stable load area, a fluctuating load area and a load transition area, and corresponding weight data is allocated to each load area type.

[0005] Optionally, in a second implementation of the first aspect of the present invention, performing a fault propagation risk assessment on grid nodes in the power grid system according to the load area type and weight data, and generating a cable capacity plan based on the risk assessment result includes: Constructing a heterogeneous propagation matrix based on the load area type and weight data, wherein the heterogeneous propagation matrix reflects the probability of a fault propagating from one grid node to an adjacent grid node; For each grid node in the power grid system, as a fault source node, multiple rounds of iterative calculations are performed using the heterogeneous propagation matrix to simulate the propagation path and propagation intensity of the fault in the power grid system, thereby obtaining a fault propagation simulation result; Calculating the fault propagation risk coefficient of each link in the power grid system under different fault scenarios based on the fault propagation simulation results; Classifying the links according to the magnitude of the fault propagation risk coefficient, and setting cable capacity redundancy coefficients for links of different risk levels; Based on the normal operating load and additional load of each link, combined with the corresponding cable capacity redundancy factor, the cable capacity specification of each link is determined and a cable capacity plan is generated.

[0006] Optionally, in a third implementation of the first aspect of the present invention, performing multiple rounds of iterative calculations on each grid node in the power grid system as a fault source node using the heterogeneous propagation matrix to simulate a fault propagation path and propagation intensity in the power grid system, and obtaining a fault propagation simulation result includes: Set the initial fault state, select the selected grid node in the current grid system as the fault source node, set the corresponding fault intensity to the maximum value, and set the fault intensity of the remaining grid nodes to zero; Calculating the fault impact intensity between the current fault source node and the adjacent grid nodes using the heterogeneous propagation matrix, and comparing the fault impact intensity of the adjacent grid nodes with a preset fault threshold; The adjacent grid nodes whose fault impact intensity exceeds the preset fault threshold are marked as new fault source nodes, and the iteration number of the new fault source node is recorded as the fault propagation delay; All newly marked fault source nodes are processed in order of topological distance from near to far, and the above propagation calculation process is repeated until no new fault source nodes are generated or the preset maximum number of iterations is reached. The fault propagation simulation results are generated based on the fault impact intensity and fault propagation delay of each grid node.

[0007] Optionally, in a fourth implementation of the first aspect of the present invention, generating a set of candidate paths based on the load area type and weight data and the cable capacity solution using a multi-objective balancing mechanism and an adaptive path growing algorithm includes: Collect spatial constraint information in the power grid system and construct a power grid environmental constraint map by combining the load area type and weight data; Establishing a multi-objective evaluation function including cost, reliability, environmental impact and adaptability, and converting the load area type and weight data into weight coefficients of each objective in the multi-objective evaluation function through a multi-objective balancing mechanism; Adopting an adaptive path growth algorithm, taking the cable starting point as the initial node, and gradually expanding the path toward the end point, in each expansion step, evaluating multiple possible expansion directions according to the multi-objective evaluation function, the power grid environment constraint map, and the cable capacity plan to obtain an evaluation result; Several expansion directions with the highest scores form branch paths, and during the path growth process, the potentially optimal path branches are retained through pruning operations; The similarities between the generated multiple paths are calculated, and paths with similarities greater than a preset threshold are screened out as a candidate path set.

[0008] Optionally, in a fifth implementation of the first aspect of the present invention, the adaptive path growth algorithm is used to gradually expand the path toward the end point with the cable start point as the initial node. In each expansion step, multiple possible expansion directions are evaluated based on the multi-objective evaluation function, the power grid environment constraint map, and the cable capacity plan. The evaluation results include: Adopting an adaptive path growth algorithm, constructing a path tree from the starting point of the cable as an initial node, and marking the initial node as a current active node; Identifying candidate extension nodes adjacent to the current active node, checking obstacles of connecting segments between the current active node and the candidate extension nodes according to the power grid environment constraint map, and calculating load adaptability in combination with the cable capacity plan; Scoring the candidate expansion nodes based on the load adaptability according to the multi-objective evaluation function, selecting the candidate expansion node with the highest comprehensive score and adding it to the path tree as a new current active node; Calculate the potential index of the expansion direction formed between each active node and prune the low-potential branches; Repeat the above expansion process until the current active node reaches the end node or reaches the preset maximum number of expansion steps, and the obtained complete path and the comprehensive score of the nodes in the complete path are used as the path evaluation result.

[0009] Optionally, in a sixth implementation of the first aspect of the present invention, performing resilience evaluation and optimization on the candidate path set under multiple preset simulation scenario conditions to generate a power grid cable routing plan includes: Calculating a path resilience index corresponding to each candidate path in the candidate path set under the preset multiple simulation scenario conditions; Calculate a comprehensive resilience score for each candidate path based on the path resilience index and the occurrence probability and impact intensity of the preset multiple simulation scenario conditions; Identify weak path segments of candidate paths whose comprehensive resilience scores are lower than a preset score threshold, and design backup paths and improve cable specifications for the weak path segments; A path sensitivity analysis is performed on the candidate path set under the preset multiple simulation scenario conditions, the stability index of the candidate path is calculated by adjusting the scenario parameters, the candidate path is selected according to the stability index, and a power grid cable wiring plan is generated.

[0010] A second aspect of the present invention provides a power grid cable wiring path optimization device, the power grid cable wiring path optimization device comprising: The load analysis module is used to perform load analysis on the geographical distribution data and power consumption data of the power grid system, and generate load area types and weight data corresponding to each area type; a risk assessment module, configured to perform a fault propagation risk assessment on grid nodes in the power grid system according to the load area type and weight data, and generate a cable capacity plan based on the risk assessment results; A path generation module, configured to generate a set of candidate paths according to the load area type and weight data and the cable capacity plan through a multi-objective balancing mechanism and an adaptive path growth algorithm; The resilience optimization module is used to evaluate and optimize the resilience of the candidate path set under multiple preset simulation scenario conditions to generate a power grid cable wiring plan.

[0011] The above-mentioned power grid cable wiring path optimization method and device, through load analysis of the geographical distribution data and electricity consumption data of the power grid system, generates load area types and weight data corresponding to each area type; based on the load area type and weight data, performs fault propagation risk assessment on the grid nodes in the power grid system, and generates a cable capacity plan based on the risk assessment results; based on the load area type and weight data and the cable capacity plan, generates a candidate path set through a multi-objective balancing mechanism and an adaptive path growth algorithm; performs resilience evaluation and optimization on the candidate path set under preset multiple simulation scenario conditions to generate a power grid cable wiring plan. The present invention comprehensively considers the load distribution characteristics of the power grid, the risk of fault propagation and the resilience of the system, and can generate a cable wiring plan that adapts to complex load environments and has fault resistance capabilities, significantly improving the reliability and stability of the power grid system.

[0012] Other features and advantages of the present invention will be described in the following description, and in part will become apparent from the description, or understood by practicing the present invention. The purposes and other advantages of the present invention are realized and obtained by the structures particularly pointed out in the description, claims and drawings.

[0013] In order to make the above-mentioned objects, features and advantages of the present invention more obvious and easy to understand, preferred embodiments are given below and described in detail with reference to the accompanying drawings. BRIEF DESCRIPTION OF THE DRAWINGS

[0014] Figure 1 This is a schematic diagram of a first embodiment of a method for optimizing a power grid cable routing path according to an embodiment of the present invention; Figure 2 Schematic diagram of an embodiment of a power grid cable wiring path optimization device in an embodiment of the present invention. DETAILED DESCRIPTION

[0015] To make the objectives, technical solutions, and advantages of the embodiments of the present invention more clear, the technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the embodiments described 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.

[0016] The terms "including," "having," and any variations thereof, as used in the embodiments of the present invention are intended to cover non-exclusive inclusions. For example, a process, method, apparatus, product, or device comprising a series of steps or units is not limited to the listed steps or units, but may optionally include other steps or units not listed, or may optionally include other steps or units inherent to the process, method, product, or device.

[0017] To facilitate understanding of this embodiment, a method for optimizing the wiring path of a power grid cable disclosed in an embodiment of the present invention is first described in detail. Figure 1 As shown, this method includes the following steps: 101. Perform load analysis on the geographical distribution data and electricity consumption data of the power grid system to generate load area types and weight data corresponding to each area type; In one embodiment of the present invention, the load analysis of the geographical distribution data and electricity consumption data of the power grid system to generate load area types and weight data corresponding to each area type includes: collecting geographical distribution data and electricity consumption data in the power grid system, wherein the geographical distribution data includes the location of the substation, the location of the distribution station and the node location of each electricity node; dividing the electricity consumption data according to time periods, and calculating the load value of each electricity node in different time periods to construct a spatiotemporal distribution matrix of the load of the electricity node; calculating the load center of gravity position in different time periods based on the spatiotemporal distribution matrix of the load of the electricity node and the geographical distribution data, and analyzing the drift trajectory of the load center of gravity position; dividing the power grid area into three load area types, namely, a stable load area, a fluctuating load area and a load transition area, based on the drift trajectory, and assigning corresponding weight data to each load area type.

[0018] Specifically, the first step is to collect geographic distribution data and electricity consumption data for the power grid system. Geographic distribution data primarily includes the locations of substations, distribution stations, and individual power consumption nodes. Substation locations are typically represented by longitude and latitude coordinates, for example, a substation is located at 116.404° longitude and 39.915° latitude. Distribution stations are also identified using longitude and latitude coordinates, with a greater number and distribution density. Power consumption nodes represent actual electricity users, such as industrial parks, commercial centers, and residential areas. This node location information forms the basic topology of the power grid system. Furthermore, the collection of electricity consumption data is crucial. This data includes actual electricity consumption records for each power consumption node at different points in time, typically spanning a year or longer to ensure that the data captures seasonal variations. This raw data is obtained through various channels, including power grid monitoring systems, smart meters, and historical electricity consumption records, forming the data foundation of the power grid system.

[0019] Specifically, after obtaining basic data, the electricity consumption data needs to be divided into time periods and load values ​​calculated. Time period division is typically based on electricity consumption characteristics, such as dividing the 24-hour cycle into peak and valley periods (e.g., 8:00 AM to 8:00 PM as the peak period, 8:00 PM to 8:00 AM the next day as the valley period), or dividing the period into four quarters based on seasons, or dividing it based on weekdays and weekends. After the division is completed, characteristic quantities such as the average load value, maximum load value, and load fluctuation amplitude are calculated for each electricity node's electricity consumption data in each time period. For example, the average load on a residential area during summer weekdays is 500 kW, the maximum load is 800 kW, and the load fluctuation amplitude is 300 kW. In this way, the load characteristics of all electricity nodes in each time period are quantified, and the results are organized into a three-dimensional matrix, the node load spatiotemporal distribution matrix. The three dimensions of this matrix are the node number, the time period number, and the load characteristic quantity type. Each element in the matrix represents a specific load characteristic value for a specific node in a specific time period.

[0020] Specifically, based on the spatiotemporal distribution matrix of the load of power consumption nodes and the geographic distribution data, the load center of gravity locations for different time periods are further calculated. The load center of gravity is the geometric center of the power grid load distribution. It is calculated by taking the weighted average of the geographic coordinates of each power consumption node and the corresponding load value. Specifically, for time period t, the longitude coordinate of the load center of gravity is calculated by multiplying the longitude coordinates of all nodes by the sum of the corresponding load values ​​divided by the total load value; the latitude coordinate is calculated similarly. In this way, a sequence of load center of gravity coordinates for each time period is obtained. Next, these coordinate sequences are subjected to time series analysis to obtain the drift trajectory of the load center of gravity over time. For example, in a certain power grid area, the load center of gravity is located near the commercial area during the day in summer and shifts toward the residential area at night. In winter, the load center of gravity generally shifts toward the industrial area. This drift trajectory intuitively reflects the spatiotemporal dynamic characteristics of the power grid load distribution.

[0021] Specifically, based on the characteristics of load center of gravity drift, power grid areas are divided into three typical load zone types. Stable load zones refer to areas where the load center of gravity remains stable for long periods of time or exhibits minimal drift, such as industrial parks where production typically continues 24 hours a day and the load is relatively stable. Fluctuating load zones refer to areas where the load center of gravity exhibits periodic and significant drift, such as commercial and office areas where the load is higher during working hours and lower during non-working hours, exhibiting significant cyclical variations. Transitional load zones are located in the transition zone between stable and fluctuating load zones, exhibiting load characteristics intermediate between the two. After the zones are divided, weights are assigned to each load zone type. These weights reflect the importance of each zone in cable routing decisions and typically include weights for cost, reliability, and environmental adaptability. For example, stable load zones may have a higher reliability weight because they typically host important production tasks; fluctuating load zones may have a higher environmental adaptability weight because their load fluctuates drastically and require greater adaptability; and transitional load zones may have intermediate weights across all dimensions.

[0022] 102. Conduct fault propagation risk assessment on grid nodes in the power grid system based on load area type and weight data, and generate a cable capacity plan based on the risk assessment results; In one embodiment of the present invention, the fault propagation risk assessment of the grid nodes in the power grid system based on the load area type and weight data, and the generation of a cable capacity plan based on the risk assessment results include: constructing a heterogeneous propagation matrix based on the load area type and weight data, wherein the heterogeneous propagation matrix reflects the probability of a fault propagating from one grid node to an adjacent grid node; for each grid node in the power grid system, as a fault source node, performing multiple rounds of iterative calculations using the heterogeneous propagation matrix to simulate the propagation path and propagation intensity of the fault in the power grid system to obtain a fault propagation simulation result; based on the fault propagation simulation result, calculating the fault propagation risk coefficient of each link in the power grid system under different fault scenarios; grading the links according to the size of the fault propagation risk coefficient, and setting a cable capacity redundancy coefficient for links of different risk levels; determining the cable capacity specification of each link based on the normal operating load and additional load borne by each link, combined with the corresponding cable capacity redundancy coefficient, to generate a cable capacity plan.

[0023] Specifically, the heterogeneous propagation matrix. This matrix, based on the previously generated load zone type and weight data, reflects the probability of a fault propagating from one grid node to an adjacent grid node. The construction process first determines the matrix dimension, which is equal to the square of the number of nodes in the grid system, forming an N×N square matrix, where N is the total number of nodes. Each element P_ij in the matrix represents the probability of a fault propagating from node i to node j. This propagation probability is influenced by multiple factors, including the physical distance between nodes, the connection type, the load zone type in which they are located, and their weights. The calculation uses a distance decay function for physical distance, where greater distances reduce the propagation probability. Regarding connection type, the propagation probability is higher for directly connected nodes, while it is lower for indirectly connected nodes. Regarding load zone type, the propagation probability between nodes is lower within stable load zones, as these zones typically employ higher-level protection measures. The propagation probability between nodes within fluctuating load zones is higher, as load fluctuations increase equipment stress. The probability is intermediate within load transition zones. Furthermore, the weights of each zone are considered. A zone with a higher weight has a lower propagation probability between nodes, reflecting the protection level of critical areas. By comprehensively considering these factors, the value of each element in the matrix is ​​calculated, thereby constructing the heterogeneous propagation matrix.

[0024] Specifically, after constructing the heterogeneous propagation matrix, the power grid fault propagation simulation phase begins. In this phase, each node in the power grid system is simulated as a fault source node, analyzing the fault propagation path and intensity. For each fault source node, the fault state vector is first initialized, with the initial fault intensity of the fault source node set to the maximum value (typically 1) and the fault intensity of all other nodes set to 0. An iterative calculation is then performed based on the heterogeneous propagation matrix, with each iteration representing a time step of fault propagation. In each iteration, the node currently in the fault state propagates the fault to its neighboring nodes according to the probability values ​​in the heterogeneous propagation matrix. Specifically, the fault intensity experienced by node j in iteration t+1 is equal to the sum of the fault intensity of all connected nodes in iteration t multiplied by the corresponding propagation probabilities. When a node's fault intensity exceeds a preset fault threshold, it is marked as an affected node, and the iteration number in which it first exceeds the threshold is recorded as the fault propagation delay. The iterative process continues until the system reaches a stable state (the change in fault intensity of all nodes is less than a preset convergence threshold) or the maximum number of iterations is reached. After completing the simulation for a fault source node, the final fault intensity and fault propagation delay for all nodes are recorded to form a set of fault propagation simulation results. This process is repeated for each node in the power grid system to obtain a complete set of fault propagation simulation results.

[0025] Specifically, based on the above-mentioned fault propagation simulation results, the fault propagation risk coefficient of each link in the power grid system is further calculated. A link refers to a physical connection that directly connects two power grid nodes and is the actual path for cable laying. The fault propagation risk coefficient reflects the importance and vulnerability of the link in the fault propagation process. The calculation process first defines the evaluation dimensions of the risk coefficient, including the propagation acceleration, propagation range, and criticality of the link. Propagation acceleration refers to the average speed at which the fault propagates through the link in all fault scenarios, and is calculated as the inverse of the propagation delay difference between two adjacent nodes; propagation range refers to the proportion of other nodes that can be affected after a fault occurs in the node connected to the link; criticality refers to the topological importance of the link in the entire network, which is usually measured by network analysis indicators such as node betweenness or connectivity. For each link, the values ​​of these three dimensions are calculated separately under different fault scenarios, and then the comprehensive risk coefficient is obtained by weighted average. For example, if a fault occurs at one end of a link within an industrial zone, the fault can quickly propagate to the other end (high propagation acceleration), but the impact is limited (low propagation range). Furthermore, the link is a critical connection within the zone (high criticality), resulting in a medium-to-high risk factor. This is how fault propagation risk factors are calculated for all links in the power grid system.

[0026] Specifically, grading the links in the power grid system according to the fault propagation risk coefficient is the basis for determining the cable capacity redundancy coefficient. The grading process uses a risk threshold method to divide the links into four levels: high risk, medium-high risk, medium risk, and low risk based on pre-set risk thresholds. For example, links with a risk coefficient greater than 0.8 are classified as high risk, links with a risk coefficient between 0.6 and 0.8 are classified as medium-high risk, links with a risk coefficient between 0.4 and 0.6 are classified as medium risk, and links with a risk coefficient less than 0.4 are classified as low risk. After the grading is completed, corresponding cable capacity redundancy coefficients are set for links of different risk levels. The cable capacity redundancy coefficient refers to the ratio of the actual configured cable capacity to the capacity required for normal operation, reflecting the system's redundancy and fault tolerance to faults. For high-risk links, a higher redundancy factor, such as 1.5, is set, meaning the actual configured cable capacity is 1.5 times the capacity required for normal operation. For medium-to-high-risk links, a medium-to-high redundancy factor, such as 1.3, is set. For medium-risk links, a medium redundancy factor, such as 1.2, is set. For low-risk links, a lower redundancy factor, such as 1.1, is set. This differentiated redundancy factor setting ensures the rational allocation of resources, focusing on protecting critical links while avoiding excessive resource waste.

[0027] Specifically, the final step is to determine the cable capacity specifications based on the load conditions and redundancy factors of each link, generating the final cable capacity plan. This step first requires calculating the normal operating load of each link, typically expressed as maximum current or maximum power. This calculation is based on the results of a load flow analysis of the power grid, taking into account the power demand of each node, the power flow distribution, and the network topology. In addition to the normal operating load, the additional load that a link may bear in the event of a failure must also be considered. This additional load primarily arises from network reconfiguration and load shifting. When certain links fail, the load they originally carried must be transferred to other links, increasing the load on these links. The magnitude of this additional load is determined through fault scenario analysis, simulating load redistribution in the event of different link failures. After determining the normal operating load and additional load, their maximum value is multiplied by the corresponding cable capacity redundancy factor to determine the required cable capacity for the link. Then, based on power system standards, standard cables are selected that meet or slightly exceed the calculated capacity. For example, a high-risk link has a normal operating load of 200A and a potential additional load of 100A. Taking the maximum value of 200A and multiplying it by the redundancy factor of 1.5 yields a required capacity of 300A. In this case, a standard cable with a rated current of 315A is selected. This way, corresponding cable specifications are determined for all links in the power grid system, forming a complete cable capacity plan.

[0028] Furthermore, the method of performing multiple rounds of iterative calculations on each grid node in the power grid system as a fault source node through the heterogeneous propagation matrix to simulate the propagation path and propagation intensity of the fault in the power grid system to obtain the fault propagation simulation result includes: setting an initial fault state, taking a selected grid node in the current power grid system as a fault source node, and setting the corresponding fault intensity to the maximum value, and setting the fault intensity of the remaining grid nodes to zero; calculating the fault impact intensity between the current fault source node and the adjacent grid nodes through the heterogeneous propagation matrix, and comparing the fault impact intensity of the adjacent grid nodes with a preset fault threshold; marking the adjacent grid nodes whose fault impact intensity exceeds the preset fault threshold as new fault source nodes, and recording the iteration number of the new fault source node as the fault propagation delay; processing all newly marked fault source nodes in order of topological distance from near to far, and repeating the above propagation calculation process until no new fault source nodes are generated or the preset maximum number of iterations is reached, and generating the fault propagation simulation result according to the fault impact intensity and fault propagation delay of each grid node.

[0029] Specifically, the key to the initial stage of power grid system fault propagation simulation lies in setting a reasonable initial fault state. This process first selects a specific grid node in the power grid system as the fault source node, such as substation A or distribution station B. This selection process can be performed sequentially through all nodes in the system or by prioritizing nodes based on their importance. For the selected fault source node, its initial fault intensity is set to the maximum value of 1.0, indicating that the node is completely faulty. Simultaneously, the initial fault intensity of all remaining grid nodes in the system is set to 0, indicating that these nodes are fully functional in their initial state. This initial state setting simulates the occurrence of a single-point fault in a real power grid, where a grid device fails first while other devices remain unaffected. The system creates and maintains a fault state vector for each node, recording the changes in the fault intensity of that node throughout the simulation. This initial fault state setting forms the starting point for the subsequent fault propagation analysis. Starting from this state, the system calculates how the fault propagates to the surrounding areas using a heterogeneous propagation matrix. For example, in a simulation, a high-voltage distribution station in the power grid system was selected as the initial fault source node, and its fault intensity was set to 1.0, indicating that the distribution station had a complete failure, while the fault intensity of the other 99 nodes in the system was 0, indicating that they had not yet been affected.

[0030] Specifically, after the initial fault state is set, the fault propagation calculation phase begins. The system uses a heterogeneous propagation matrix to calculate the fault impact intensity between the current fault source node and its neighboring grid nodes. Neighboring nodes here refer to all nodes directly connected to the fault source node, including substations, distribution stations, or power terminals directly connected via transmission lines, distribution lines, or other power connections. During the calculation, all elements in the heterogeneous propagation matrix corresponding to the fault source node are extracted. Each element value represents the probability of the fault propagating from the fault source node to a specific neighboring node. The current fault intensity of the fault source node is then multiplied by these probabilities to calculate the fault impact intensity for each neighboring node. For example, if the current fault intensity of a fault source node is 1.0, and according to the heterogeneous propagation matrix, the probabilities of propagation from this node to three neighboring nodes are 0.7, 0.5, and 0.3, respectively, then the fault impact intensities on these three neighboring nodes are 0.7, 0.5, and 0.3, respectively. After calculating the fault impact intensities of the neighboring nodes, the system compares these values ​​with preset fault thresholds. The fault threshold is a criterion for determining whether a node is significantly impacted. It is typically set between 0.2 and 0.5, depending on the grid system's protection capabilities and fault tolerance. If the impact of a fault on an adjacent node is less than the threshold, the node is considered able to withstand the impact without experiencing significant failure. If the impact is greater than or equal to the threshold, the node is considered significantly impacted and at risk of failure.

[0031] Specifically, for adjacent nodes whose fault impact intensity exceeds a preset threshold, the system will mark them as new fault source nodes. This step reflects the cascading failure phenomenon in the power grid system, that is, the failure of one node may cause the adjacent nodes to also fail, thereby forming a fault spread. During the marking process, the system will update the fault intensity values ​​of these new fault source nodes, usually taking the larger value of the calculated fault impact intensity and the current fault intensity to reflect the cumulative effect. At the same time, the system will also record the iteration number of each new fault source node as a quantitative indicator of the fault propagation delay. The fault propagation delay reflects the time step required for the fault to propagate from the initial fault source node to the current node, and is an important parameter for evaluating the fault propagation speed.

[0032] Specifically, after marking new fault source nodes, the system needs to determine the order in which to process them so that fault propagation can continue in the next iteration. This processing order is determined based on the principle of topological distance, from closest to furthest. This prioritizes new fault source nodes that are topologically closer to the initial fault source node. Topological distance refers to the number of edges contained in the shortest path between two nodes and reflects the connectivity between nodes in the power grid. This processing order is based on the physical characteristics of power grid fault propagation. Typically, a fault first affects devices that are physically close or directly connected, and then gradually spreads to more distant areas. In implementation, the system calculates the topological distance between all newly marked fault source nodes and the initial fault source node and then arranges these nodes into a processing queue in ascending order of distance. In the next iteration, the system, following this queue order, calculates the fault impact on adjacent nodes with each new fault source node as the center, and repeats the aforementioned propagation calculation process. For example, in a simulation, five new fault source nodes are marked, and their topological distances from the initial fault source node are 1, 1, 2, 2, and 3, respectively. The system will first process the two nodes with a distance of 1, then the two nodes with a distance of 2, and finally the node with a distance of 3.

[0033] Specifically, fault propagation simulation is an iterative process, and the system will continuously repeat the above calculation steps until the termination condition is met. There are two main termination conditions: one is that no new fault source nodes are generated, indicating that the fault has stopped spreading; the other is that the preset maximum number of iterations is reached, which is usually set to about twice the size of the power grid system to ensure that the fault can be fully propagated in most cases. During the iteration process, the system will continuously update the fault intensity value of each node, record the marking status of the fault source node and the fault propagation delay. When the termination condition is met, the system generates a complete fault propagation simulation result based on the final node fault intensity and fault propagation delay. This result usually includes two parts of information: one is the final fault intensity of each node, which reflects the degree of impact on the node in the current fault scenario; the other is the fault propagation delay of each node, which reflects the time dynamic characteristics of the fault propagation. This information constitutes the basic data for assessing the risk of link fault propagation. For example, the final results of a certain simulation showed that among a total of 100 nodes, the fault intensity of 35 nodes exceeded the threshold and became fault source nodes. Their fault propagation delays ranged from 1 to 8, with an average of 3.2, indicating that the average propagation time from the initial fault source node to these nodes was 3.2 time steps.

[0034] 103. Based on the load area type and weight data and the cable capacity plan, a set of candidate paths is generated through a multi-objective balancing mechanism and an adaptive path growing algorithm; In one embodiment of the present invention, the generation of a candidate path set based on the load area type and weight data and the cable capacity plan through a multi-objective balancing mechanism and an adaptive path growth algorithm includes: collecting spatial constraint information in the power grid system, and constructing a power grid environment constraint map in combination with the load area type and weight data; establishing a multi-objective evaluation function including cost, reliability, environmental impact and adaptability, and converting the load area type and weight data into weight coefficients of each objective in the multi-objective evaluation function through a multi-objective balancing mechanism; using an adaptive path growth algorithm, with the cable starting point as the initial node, gradually expanding the path toward the end point, and in each expansion step, evaluating multiple possible expansion directions according to the multi-objective evaluation function, the power grid environment constraint map and the cable capacity plan to obtain an evaluation result; several expansion directions with the highest scores form branch paths, and in the path growth process, the potentially optimal path branches are retained through pruning operations; calculating the similarity between the multiple generated paths, and screening out paths with a similarity greater than a preset threshold as the candidate path set.

[0035] Specifically, the first step in optimizing cable routing is to collect spatial constraint information within the power grid system and construct a grid environmental constraint map. Spatial constraint information includes factors such as terrain obstacles, building distribution, road locations, rivers and lakes, protected areas, and existing infrastructure. This information is typically obtained from geographic information system (GIS) databases, satellite imagery, topographic maps, and field survey records. For example, slope data in mountainous areas, building outlines in urban areas, the precise location of road networks, and the width and direction of rivers can be obtained. After obtaining this raw data, it is integrated with the previously mentioned load area type and weight data to construct a grid environmental constraint map. This is achieved by dividing the entire planning area into grid cells, each containing location coordinates and constraint attributes. Constraint attributes are categorized as hard and soft constraints. Hard constraints represent areas where cables absolutely cannot traverse, such as building foundations and core areas of nature reserves. These areas are marked as no-go zones in the constraint map. Soft constraints represent areas where cable traversal would increase cost or risk, such as rivers, roads, and slopes. These areas are assigned different weights in the constraint map based on the difficulty of traversal. At the same time, load area type information is overlaid on the constraint map, with stable load areas, fluctuating load areas, and load transition areas each distinguished by a unique identifier. Weight data for each area is also integrated into the attributes of the corresponding grid cell. The resulting grid environmental constraint map is a multi-layered spatial data structure that clearly expresses the various constraints and regional characteristics that need to be considered during cable routing.

[0036] Specifically, based on the constructed grid environmental constraint map, the next step is to establish a multi-objective evaluation function and determine the weight coefficients for each objective through a multi-objective balancing mechanism. This multi-objective evaluation function typically includes four core indicators: cost, reliability, environmental impact, and adaptability. The cost indicator primarily considers the sum of cable material costs, construction costs, and operation and maintenance costs, with cable length being the primary influencing factor. The reliability indicator reflects the ability of a cable line to maintain normal operation under various conditions and is closely related to cable specifications, wiring methods, and the environment along the line. The environmental impact indicator measures the degree of interference of cable wiring with the natural environment and human activities, including ecological damage, landscape impact, and electromagnetic radiation. The adaptability indicator assesses the ability of the line to adapt to future load changes and network expansion and is related to line margin and topology. These four indicators have different dimensions and magnitudes and require normalization to bring their values ​​into the range [0, 1]. Then, through the multi-objective balancing mechanism, the aforementioned load area types and weight data are converted into weight coefficients for each indicator in the multi-objective evaluation function. The conversion process follows the following principles: In stable load areas, reliability indicators are weighted higher because such areas typically carry critical loads; in fluctuating load areas, adaptability indicators are weighted higher to cope with frequent load changes; in load transition areas, cost and environmental impact indicators are weighted relatively higher because these areas typically have greater wiring flexibility. The specific conversion uses a weighted mapping method to calculate the final indicator weight coefficient based on the area ratio and weight intensity of each area type. For example, in a certain planning area, the stable load area accounts for 40%, the fluctuating load area accounts for 35%, and the load transition area accounts for 25%. The weight coefficients of the four indicators may be: cost 0.25, reliability 0.35, environmental impact 0.15, and adaptability 0.25, reflecting the high reliability requirements of this area.

[0037] Specifically, after determining the multi-objective evaluation function, an adaptive path growing algorithm is used for cable path planning. This algorithm begins with the cable's starting point and gradually grows the path toward the destination through iterative expansion. The starting point is typically a power source or substation, while the destination is a load center or target distribution station. The path growing process is an iterative one. In each expansion step, the algorithm first determines the current active node, the most advanced node in the path. For the current active node, the algorithm identifies all possible expansion directions within a certain radius around it, typically at angular intervals of 15 or 30 degrees, generating 12 to 24 candidate directions. For each candidate direction, the algorithm performs the following checks and evaluations: First, the algorithm checks whether the line segment extending along that direction intersects the hard constraint area in the grid environmental constraint map. If so, the direction is excluded. Second, the algorithm checks whether the expansion in that direction complies with the cable capacity plan requirements, ensuring that the path can accommodate the planned cable specifications. Finally, the algorithm calculates the expansion score for that direction based on the multi-objective evaluation function. The scoring process includes: calculating a cost score for the expansion segment, primarily based on length and the difficulty of traversing the area; calculating a reliability score, taking into account risk factors along the segment; calculating an environmental impact score, assessing the degree of disruption to the surrounding environment; and calculating an adaptability score, analyzing the segment's connectivity with the endpoint and potential for future expansion. Each score is weighted and summed using the aforementioned weighting coefficients to produce a comprehensive score for the expansion direction. In this way, the algorithm comprehensively evaluates all feasible expansion directions and generates an assessment result.

[0038] Specifically, after obtaining the evaluation results for each expansion direction, the algorithm selects the highest-scoring directions (typically the top three to five) for actual expansion, forming multiple branching paths. Each branching path grows independently, with its own current active node and historical trajectory. To control computational complexity and avoid inefficient searches, the algorithm implements pruning during the path growth process, retaining potentially optimal path branches. This pruning strategy is based on two core metrics: the cumulative score of the current path and a heuristic estimate of the distance to the endpoint. The cumulative score is the weighted sum of the scores of all expansion steps on the path, reflecting the quality of the path's already traveled portion. The heuristic estimate predicts the potential cost and quality of the remaining path portion, typically based on the straight-line distance from the current node to the endpoint and adjusted based on information from the power grid environmental constraint map. These two metrics are combined into a potential index, which is used to rank all active path branches. The top N branches with the highest potential index are retained (N is typically set to half or one-third of the total number of branches), while branches with lower potential are removed. For example, if there are 12 active branches after a certain iteration, the algorithm calculates the potential index of each branch, retains the four branches with the highest index for continued growth, and removes the remaining eight branches. The pruning operation significantly improves the efficiency of the algorithm, concentrating computing resources on the most promising path direction. At the same time, it maintains the diversity of the search by retaining multiple branches and avoids the trap of local optimal solutions.

[0039] Specifically, when the adaptive path growing algorithm completes an iteration (usually when it reaches the endpoint or the maximum number of iterations), multiple complete paths from the starting point to the end point are obtained. These paths typically number between 5 and 20, each with different characteristics and advantages. Computing the similarity between these paths ensures diversity in the candidate set and avoids selecting redundant paths that are too similar. Path similarity is calculated using a spatial overlap method, representing each path as a set of grid cells. The ratio of the number of grid cells shared by the two paths to the total number of grid cells is then calculated. For example, if path A passes through the grid cell set {1, 2, 3, 5, 7, 9} and path B passes through the grid cell set {1, 2, 4, 6, 8, 9}, and both paths share the grid cell {1, 2, 9}, totaling the grid cells {1, 2, 3, 4, 5, 6, 7, 8, 9}, then the similarity is 3 / 9 = 0.33, indicating a 33% overlap between the two paths. After calculating the similarity between all path pairs, a cluster analysis method is used to group paths with a similarity above a preset threshold (typically set at 0.7 or 0.8) into the same cluster. Within each cluster, the path with the highest overall score is selected to represent that cluster. For example, if 15 paths were originally generated and grouped into four clusters after similarity analysis, the highest-scoring path from each cluster is selected, resulting in four significantly different candidate paths that form the final set of candidate paths. This approach ensures both the quality of path selection (selecting the best path within each cluster) and the diversity of the candidate set.

[0040] Furthermore, the adaptive path growth algorithm is used, with the cable starting point as the initial node, to gradually expand the path toward the end point. In each expansion step, multiple possible expansion directions are evaluated according to the multi-objective evaluation function, the power grid environment constraint map and the cable capacity plan, and the evaluation results include: using the adaptive path growth algorithm, constructing a path tree from the cable starting point as the initial node, and marking the initial node as the current active node; identifying candidate expansion nodes adjacent to the current active node, checking the obstacle conditions of the connecting line segments between the current active node and the candidate expansion nodes according to the power grid environment constraint map, and calculating the load adaptability in combination with the cable capacity plan; scoring the candidate expansion nodes based on the load adaptability according to the multi-objective evaluation function, selecting the candidate expansion node with the highest comprehensive score and adding it to the path tree as the new current active node; calculating the potential index of the expansion direction formed between each active node and pruning low-potential branches; repeating the above expansion process until the current active node reaches the end node or reaches the preset maximum number of expansion steps, and taking the obtained complete path and the comprehensive score of the nodes in the complete path as the path evaluation result.

[0041] Specifically, the adaptive path growth algorithm begins by constructing a path tree with the cable's starting point as the initial node. A path tree is a specialized data structure used to record and track multiple possible paths from a starting point to a destination. The construction process first requires determining the precise location of the cable's starting point, typically a substation or key distribution node on the power supply side, such as a 500kV main substation in the eastern part of a city or a 110kV distribution station in an industrial park. Once the starting point is determined, a root node is created in the path tree and marked as the currently active node. The active node is the leading node in the current path expansion and serves as the reference point for the algorithm to determine the next expansion direction. When creating the root node, the system initializes node attributes, including the node's spatial coordinates, cumulative score (initial value 0), current load zone type, path length to date (initial value 0), and connectivity with the starting point and destination. In addition to creating the root node, the system also initializes overall path tree properties, such as the maximum number of branches, maximum iteration depth, and termination criteria. The path tree data structure is represented using an adjacency list or adjacency matrix to efficiently add new nodes and record connectivity between nodes. For example, in a certain planning process, the starting point of the cable is a 220kV substation in the north of the city with the geographic coordinates of (40.02°N, 116.34°E). The system creates an initial node, marks it as the current active node, sets the maximum number of branches to 8, the maximum iteration depth to 200, and the termination condition is reaching the end point or exceeding the maximum iteration depth.

[0042] Specifically, after the initial node of the path tree is established, the algorithm enters the iterative expansion phase, first identifying candidate expansion nodes around the current active node. Candidate expansion nodes are the next locations that can be directly reached from the current active node. The identification process takes into account the actual constraints of power grid planning. In specific implementation, a series of candidate points are generated along different directions within a predefined search radius (typically the length of a cable unit segment, such as 100 meters or 500 meters), centered on the current active node. A common method is uniform angle sampling, which generates candidate points in 12 or 24 directions within a 360-degree range at intervals of 15° or 30°. For each candidate expansion node, the algorithm checks whether the connecting line segment from the current active node to the candidate node intersects with obstacles in the power grid environmental constraint map. Obstacles include hard-constrained areas such as buildings, water bodies, and protected areas, which are clearly identified in the environmental constraint map. This check uses a line segment-polygon intersection test algorithm. If the connecting line segment intersects any obstacle polygon, the candidate expansion node is excluded. For example, a currently active node is located in a city center, surrounded by multiple commercial buildings. The algorithm generates candidate expansion nodes in 24 directions around it. After checking for obstacles, seven of these candidates are eliminated due to intersecting buildings. For those candidate expansion nodes that pass the obstacle check, the algorithm further calculates load suitability based on the cable capacity plan. Load suitability reflects the degree to which the cable capacity requirements at the location of the connection segment match the planned capacity. The calculation method maps the connection segment to a link in the power grid system, obtains the link specifications from the cable capacity plan, and then evaluates whether the current connection location is suitable for laying cables of that specification. Suitability values ​​typically range from 0 to 1, with higher values ​​indicating greater suitability for laying cables of the planned specifications. For example, a connection segment located in an area with flat terrain and good soil conditions is suitable for laying cables of any specification, with a load suitability of 0.95. However, another connection segment, which crosses a rocky area, is only suitable for laying small cables, despite the planned requirement for large-capacity cables. Its load suitability is only 0.3.

[0043] Specifically, after determining the obstacle conditions and load adaptability of candidate expansion nodes, the algorithm comprehensively scores each valid candidate expansion node based on a multi-objective evaluation function. The scoring process comprehensively considers four core metrics: cost, reliability, environmental impact, and adaptability. The cost metric primarily considers the length of the connecting segment and the terrain complexity of the traversed area. Shorter lengths and simpler terrain result in higher cost scores. The reliability metric considers the safety and stability of the connecting segment's environment, including factors such as natural disaster risk and the potential for human interference. A safer and more stable environment results in a higher reliability score. The environmental impact metric assesses the extent of the connecting segment's interference with the surrounding ecological environment and human activities, including vegetation damage, landscape impact, and electromagnetic radiation. Minimum interference results in a higher environmental impact score. The adaptability metric considers the connecting segment's ability to adapt to future load changes and network expansion, as well as its proximity to the endpoint. Stronger adaptability and closer proximity to the endpoint result in a higher adaptability score. The previously calculated load adaptability serves as a key input in the scoring process, directly influencing the scores for both the reliability and adaptability metrics. The scores for each metric are weighted and summed using the weight coefficients in the multi-objective evaluation function to produce a comprehensive score for the candidate expansion node. For example, a candidate expansion node has four metrics scores of 0.8 for cost, 0.7 for reliability, 0.9 for environmental impact, and 0.6 for adaptability. The corresponding weight coefficients are 0.3, 0.3, 0.2, and 0.2, resulting in a comprehensive score of 0.8 × 0.3 + 0.7 × 0.3 + 0.9 × 0.2 + 0.6 × 0.2 = 0.75. The algorithm calculates the comprehensive score for all valid candidate expansion nodes and selects the node with the highest score, adding it to the path tree as the new current active node. For example, in one iteration, the current active node has five valid candidate expansion nodes with comprehensive scores of 0.75, 0.68, 0.82, 0.71, and 0.63, respectively. The algorithm selects the candidate node with a score of 0.82, adds it to the path tree, and updates it as the new current active node.

[0044] Specifically, as the path tree continues to expand, the number of accumulated nodes and branches rapidly increases. To control computational complexity, the algorithm calculates the potential index of each active branch and performs pruning operations. The potential index is a comprehensive metric for assessing the future development potential of a path branch. Its calculation method comprehensively considers the current branch's cumulative score, its expanded length, and its distance from the endpoint. The cumulative score is a weighted average of the scores of all nodes on the branch, reflecting the path quality. The expanded length is the actual path length from the branch's starting point to the current active node, reflecting the path's physical characteristics. The distance from the endpoint is expressed as the inverse of the distance, with closer distances indicating a greater contribution. These three components are combined using preset weights to create the final potential index. A higher index value indicates a higher likelihood of the branch developing into a high-quality path. For example, a branch with a cumulative score of 0.78, an expanded length of 2.5 kilometers, and a straight-line distance of 1.2 kilometers from the current active node to the endpoint has a calculated potential index of 0.85, placing it at the top of all active branches. After calculating the potential index of all active branches, the algorithm sorts the branches from highest to lowest by index value, retaining the top N branches (N is typically half or two-thirds of the preset maximum number of branches) and removing the remaining low-potential branches. This pruning operation significantly reduces the number of branches to be processed, concentrating computational resources on high-potential branches and improving algorithm efficiency. For example, in a case where there were 32 active branches before pruning, the algorithm retained the top 12 high-potential branches after calculating the potential index, while discarding the remaining 20 low-potential branches. This reduced the computational effort by 62.5%, with minimal impact on the quality of the final result.

[0045] Specifically, the path growth algorithm repeatedly executes the above expansion process, continuously advancing the path tree toward the destination. Each iteration includes identifying candidate expansion nodes, checking for obstacles, calculating load adaptability, performing a comprehensive scoring, selecting the optimal node to add to the path tree, calculating a potential index, and performing pruning. The iterative process continues until a termination condition is met. There are two main termination conditions: the current active node reaches the destination node, and the preset maximum number of expansion steps is reached. Reaching the destination is typically determined using a distance threshold: when the distance between the current active node and the destination node is less than a preset threshold (such as 50 or 100 meters), the destination is considered reached. The maximum number of expansion steps is set to prevent the algorithm from looping indefinitely if a valid path cannot be found. It is typically set to a number of steps corresponding to 1.5 to 2 times the expected path length. When the current active node of a branch reaches the destination, the algorithm records the complete path corresponding to that branch, including the sequence of all nodes along the path and their attribute information. The comprehensive score of the complete path is a weighted average of the scores of all nodes along the path, reflecting the overall quality of the entire path. The algorithm continues to execute until all active branches reach the end point or the maximum number of expansion steps is reached, or there are no remaining active branches (all pruned). Finally, the algorithm outputs all complete paths that reach the end point and their comprehensive scores as the path evaluation results.

[0046] 104. Under the preset conditions of multiple simulation scenarios, the candidate path set is evaluated and optimized for resilience, and a grid cable wiring plan is generated.

[0047] In one embodiment of the present invention, the resilience evaluation and optimization of the candidate path set under the preset multiple simulation scenario conditions to generate a power grid cable wiring plan includes: calculating the path resilience index corresponding to each candidate path in the candidate path set under the preset multiple simulation scenario conditions; calculating the comprehensive resilience score of each candidate path based on the path resilience index, combined with the occurrence probability and impact intensity in the preset multiple simulation scenario conditions; identifying the weak path segments of the candidate paths whose comprehensive resilience scores are lower than the preset score threshold, and performing backup path design and cable specification improvement on the weak path segments; performing path sensitivity analysis on the candidate path set under the preset multiple simulation scenario conditions, calculating the stability index of the candidate paths by adjusting the scenario parameters, selecting the candidate paths according to the stability index, and generating a power grid cable wiring plan.

[0048] Specifically, the resilience evaluation of a power grid cable routing scheme first requires calculating the path resilience index of each candidate path under a variety of preset simulation scenarios. These simulation scenarios are pre-set based on the various extreme situations that the power grid system may face, and typically include natural disaster scenarios (such as earthquakes, floods, and typhoons), human interference scenarios (such as construction excavation and malicious sabotage), equipment failure scenarios (such as cable aging and connector failures), and load mutation scenarios (such as peak power consumption and regional load surges). Each scenario has a clear description and numerical parameters. For example, an earthquake scenario is defined as a magnitude 6.0 earthquake with an epicenter in the northeast corner of the planning area and an impact radius of 30 kilometers; a flood scenario is defined as a once-in-a-century flood, with the inundation area covering a 5-kilometer radius along the river bank in the southwest of the planning area. For each path in the candidate path set, the system places it in these preset scenarios and simulates the path's performance under the conditions of each scenario. The calculation of the path resilience index includes three dimensions: fault resistance, recovery speed, and service maintenance capability. Fault resilience measures the extent to which a path maintains physical integrity under scenario impacts. This is calculated by counting the proportion of nodes on the path affected by the scenario; the smaller the impact, the higher the score. Recovery speed reflects how long it takes for a path to return to normal function after damage. This is related to factors such as path accessibility and material supply difficulties; faster recovery times result in higher scores. Service maintenance capability assesses the ability of a path to maintain basic services even in the event of partial damage. This is related to path redundancy and network topology; stronger maintenance capability results in higher scores. These three metrics are weighted averaged to form the final path resilience index.

[0049] Specifically, based on the path resilience index matrix, the next step is to calculate a comprehensive resilience score for each candidate path. This comprehensive resilience score takes into account the path's performance under various scenarios and the characteristics of the scenarios themselves, reflecting the path's overall resilience. The calculation process first considers two key parameters: the probability of occurrence and the impact intensity of each simulated scenario. The probability of occurrence reflects the likelihood of a scenario occurring and is typically determined based on historical data and expert assessments. For example, the annual probability of an earthquake scenario is 0.5%, and the annual probability of a flood scenario is 2%. The impact intensity reflects the impact of the scenario on the system if it occurs, typically using a quantitative scale of 1-10, such as an impact intensity of 8 for an earthquake scenario and 6 for a flood scenario. Multiplying these two parameters yields the scenario's risk weight. Scenarios with higher risk weights have a greater impact in the comprehensive score. The comprehensive resilience score is calculated by multiplying the path's resilience index under each scenario by the corresponding scenario's risk weight, then normalizing the score. Specifically, for candidate path i, its comprehensive resilience score is the sum of the resilience index under all scenario j multiplied by the scenario's risk weight, divided by the sum of the risk weights. This calculated comprehensive resilience score comprehensively reflects the path's overall performance in the face of various risks.

[0050] Specifically, for candidate paths whose comprehensive resilience scores fall below a preset threshold, weak path segments are identified and targeted for optimization. Weak path segments are those portions of a path that are prone to failure or severe impact under simulated scenario conditions and are key to improving path resilience. The identification process first analyzes the performance of each node and connecting segment on the path under different scenarios, calculating node and segment vulnerabilities. Node vulnerability is a weighted average of the node's impact across all scenarios, reflecting the vulnerability of the node's location. Segment vulnerability takes into account the vulnerabilities of the nodes at both ends of the segment as well as the physical characteristics of the segment itself (such as length and the type of area it traverses), reflecting the segment's vulnerability. Nodes and segments are sorted by vulnerability, and a threshold is set (typically 1.5 times the average vulnerability). Nodes and segments with vulnerabilities exceeding the threshold are marked as weak points. Connectivity analysis then merges adjacent weak points into weak path segments. For example, if the vulnerability of five nodes on a candidate path exceeds the threshold, located in the 7th-8th, 15th-17th, and 23rd sections of the path, three weak path sections are identified. For the identified weak path sections, two main optimization strategies are implemented: backup path design and cable specification improvement. Backup path design is to plan additional backup paths near the weak sections, which can be quickly switched to when the main path is damaged. In specific implementation, a detour start and end point are set for each weak section, and the path planning algorithm is re-run between the two points to generate a backup path that is sufficiently separated from the original path (usually a spatial distance of no less than 100 meters is required). Cable specification improvement is to use higher-specification cables or strengthen protection measures for weak sections, such as increasing the number of cable sheath layers, increasing the cable cross-section, and using waterproof and fire-proof materials. The combination of these two strategies significantly improves the resilience level of weak path sections, thereby improving the overall performance of the entire path.

[0051] Specifically, after optimization is complete, a sensitivity analysis is performed on the candidate paths to assess their stability under uncertain conditions. Sensitivity analysis involves adjusting simulation scenario parameters and observing changes in the path resilience index to determine the path's adaptability to environmental changes. In practice, key parameters of the original pre-set scenario are slightly fluctuated, typically within a ±20% range, to generate multiple perturbation scenarios. For example, if the original earthquake scenario had a magnitude of 6.0 and an impact radius of 30 kilometers, the perturbation generates multiple scenarios with magnitudes ranging from 4.8 to 7.2 and impact radiuses ranging from 24 to 36 kilometers. The resilience index of the candidate paths is recalculated under these perturbation scenarios and compared with the resilience index under the original scenario to calculate the rate of change. A smaller rate of change indicates a less sensitive path to environmental parameter changes and greater stability. The results from all perturbation scenarios are combined to calculate a stability index for each candidate path. The stability index is typically expressed as the inverse of the average rate of change of the resilience index, with higher values ​​indicating a more stable path. For example, if the average change rate of the resilience index of a candidate path under five disturbance scenarios is 8%, its stability index is 1 / 0.08 = 12.5. The final selection of candidate paths is based on the stability index and the aforementioned comprehensive resilience score. The selection process considers both the path's resilience level and its stability performance. A weighted ranking method is typically used. This involves combining the comprehensive resilience score and stability index according to preset weights (such as 0.7 and 0.3) to create a final score. The path with the highest score is then selected as the recommended option. In certain circumstances, multiple paths with strong complementarity are combined to form a network layout, further enhancing the overall resilience of the system. For example, a planning project ultimately selected the two paths ranked first and third in comprehensive scores (the second and third paths were too similar), forming a ring structure that significantly improved the system's risk tolerance. The final selected path or path combination, together with detailed information such as path coordinates, node attributes, cable specifications, and protection measures, constitutes a complete power grid cable routing plan.

[0052] In this embodiment, load analysis is performed on the geographical distribution data and electricity consumption data of the power grid system to generate load area types and weight data corresponding to each area type; based on the load area types and weight data, a fault propagation risk assessment is performed on the grid nodes in the power grid system, and a cable capacity plan is generated based on the risk assessment results; based on the load area types and weight data and the cable capacity plan, a candidate path set is generated through a multi-objective balancing mechanism and an adaptive path growth algorithm; the candidate path set is evaluated and optimized for resilience under various preset simulation scenarios to generate a power grid cable wiring plan. The present invention comprehensively considers the load distribution characteristics of the power grid, the fault propagation risk, and the system resilience, and can generate a cable wiring plan that adapts to complex load environments and has fault resistance capabilities, significantly improving the reliability and stability of the power grid system.

[0053] The above describes the method for optimizing the wiring path of power grid cables according to the embodiment of the present invention. The following describes the device for optimizing the wiring path of power grid cables according to the embodiment of the present invention. Figure 2 In one embodiment of the present invention, a device for optimizing a power grid cable routing path includes: The load analysis module 201 is used to perform load analysis on the geographical distribution data and power consumption data of the power grid system, and generate load area types and weight data corresponding to each area type; a risk assessment module 202 for performing a fault propagation risk assessment on grid nodes in the power grid system according to the load area type and weight data, and generating a cable capacity plan based on the risk assessment results; A path generation module 203 is configured to generate a set of candidate paths based on the load area type and weight data and the cable capacity plan through a multi-objective balancing mechanism and an adaptive path growing algorithm; The resilience optimization module 204 is used to evaluate and optimize the resilience of the candidate path set under various preset simulation scenario conditions to generate a power grid cable wiring plan.

[0054] In an embodiment of the present invention, the power grid cable wiring path optimization device runs the above-mentioned power grid cable wiring path optimization method, and the power grid cable wiring path optimization device generates load area types and weight data corresponding to each area type by performing load analysis on the geographical distribution data and power consumption data of the power grid system; performs fault propagation risk assessment on the power grid nodes in the power grid system according to the load area type and weight data, and generates a cable capacity plan according to the risk assessment result; generates a candidate path set through a multi-objective balancing mechanism and an adaptive path growth algorithm according to the load area type and weight data and the cable capacity plan; performs resilience evaluation and optimization on the candidate path set under preset multiple simulation scenario conditions to generate a power grid cable wiring plan. The present invention comprehensively considers the load distribution characteristics of the power grid, the risk of fault propagation and the resilience of the system, and can generate a cable wiring plan that adapts to complex load environments and has fault resistance capabilities, thereby significantly improving the reliability and stability of the power grid system.

[0055] Those skilled in the art will clearly understand that, for the convenience and brevity of description, the specific working processes of the above-described systems, devices, and units can refer to the corresponding processes in the aforementioned method embodiments and will not be repeated here.

[0056] If the integrated unit is implemented as 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 portion that contributes to the prior art, or all or part of the technical solution, can be embodied in the form of a software product. This 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 execute all or part of the steps of the method described in each embodiment of the present invention. The aforementioned storage medium includes various media that can store program code, such as a USB flash drive, a mobile hard drive, a read-only memory (ROM), a random access memory (RAM), a magnetic disk, or an optical disk.

[0057] As described above, the above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit the same. Although the present invention has been described in detail with reference to the above embodiments, those skilled in the art should understand that the technical solutions described in the above embodiments can still be modified, or some of the technical features thereof can be replaced by equivalents. However, these modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A method for optimizing power grid cable routing, characterized in that: The grid cable routing path optimization method comprises: Perform load analysis on the geographical distribution data and electricity consumption data of the power grid system to generate load area types and weight data corresponding to each area type; performing a fault propagation risk assessment on grid nodes in the power grid system based on the load area type and weight data, and generating a cable capacity plan based on the risk assessment results; Generate a set of candidate paths based on the load area type and weight data and the cable capacity plan through a multi-objective balancing mechanism and an adaptive path growing algorithm; The candidate path set is evaluated for resilience and optimized under various preset simulation scenarios to generate a power grid cable wiring plan.

2. The method for optimizing the wiring path of power grid cables according to claim 1, characterized in that: The load analysis of the geographical distribution data and power consumption data of the power grid system to generate load area types and weight data corresponding to each area type includes: Collecting geographical distribution data and electricity consumption data in the power grid system, wherein the geographical distribution data includes the location of substations, the location of distribution stations, and the node location of each electricity consumption node; Dividing the power consumption data into time periods, calculating the load value of each power consumption node in different time periods, and constructing a spatiotemporal distribution matrix of the power consumption node load; Calculating the load center of gravity position in different time periods based on the spatiotemporal distribution matrix of the power node load and the geographical distribution data, and analyzing the drift trajectory of the load center of gravity position; According to the drift trajectory, the power grid area is divided into three load area types: a stable load area, a fluctuating load area and a load transition area, and corresponding weight data is allocated to each load area type.

3. The method for optimizing power grid cable routing according to claim 1, wherein: The performing of a fault propagation risk assessment on the grid nodes in the power grid system according to the load area type and weight data, and generating a cable capacity plan according to the risk assessment result includes: Constructing a heterogeneous propagation matrix based on the load area type and weight data, wherein the heterogeneous propagation matrix reflects the probability of a fault propagating from one grid node to an adjacent grid node; For each grid node in the power grid system, as a fault source node, multiple rounds of iterative calculations are performed using the heterogeneous propagation matrix to simulate the propagation path and propagation intensity of the fault in the power grid system, thereby obtaining a fault propagation simulation result; Calculating the fault propagation risk coefficient of each link in the power grid system under different fault scenarios based on the fault propagation simulation results; Classifying the links according to the magnitude of the fault propagation risk coefficient, and setting cable capacity redundancy coefficients for links of different risk levels; Based on the normal operating load and additional load of each link, combined with the corresponding cable capacity redundancy factor, the cable capacity specification of each link is determined and a cable capacity plan is generated.

4. The method for optimizing power grid cable routing according to claim 3, wherein: The method of performing multiple rounds of iterative calculations on each grid node in the power grid system as a fault source node using the heterogeneous propagation matrix to simulate the propagation path and propagation intensity of the fault in the power grid system and obtain the fault propagation simulation results includes: Set the initial fault state, select the selected grid node in the current grid system as the fault source node, set the corresponding fault intensity to the maximum value, and set the fault intensity of the remaining grid nodes to zero; Calculating the fault impact intensity between the current fault source node and the adjacent grid nodes using the heterogeneous propagation matrix, and comparing the fault impact intensity of the adjacent grid nodes with a preset fault threshold; The adjacent grid nodes whose fault impact intensity exceeds the preset fault threshold are marked as new fault source nodes, and the iteration number of the new fault source node is recorded as the fault propagation delay; All newly marked fault source nodes are processed in order of topological distance from near to far, and the above propagation calculation process is repeated until no new fault source nodes are generated or the preset maximum number of iterations is reached. The fault propagation simulation results are generated based on the fault impact intensity and fault propagation delay of each grid node.

5. The method for optimizing power grid cable routing according to claim 1, wherein: Generating a set of candidate paths according to the load area type and weight data and the cable capacity plan through a multi-objective balancing mechanism and an adaptive path growing algorithm includes: Collect spatial constraint information in the power grid system and construct a power grid environmental constraint map by combining the load area type and weight data; Establishing a multi-objective evaluation function including cost, reliability, environmental impact and adaptability, and converting the load area type and weight data into weight coefficients of each objective in the multi-objective evaluation function through a multi-objective balancing mechanism; Adopting an adaptive path growth algorithm, taking the cable starting point as the initial node, and gradually expanding the path toward the end point, in each expansion step, evaluating multiple possible expansion directions according to the multi-objective evaluation function, the power grid environment constraint map, and the cable capacity plan to obtain an evaluation result; Several expansion directions with the highest scores form branch paths, and during the path growth process, the potentially optimal path branches are retained through pruning operations; The similarities between the generated multiple paths are calculated, and paths with similarities greater than a preset threshold are screened out as a candidate path set.

6. The method for optimizing the wiring path of power grid cables according to claim 5, characterized in that: The adaptive path growth algorithm is used to gradually expand the path toward the end point with the cable starting point as the initial node. In each expansion step, multiple possible expansion directions are evaluated based on the multi-objective evaluation function, the power grid environment constraint map, and the cable capacity plan. The evaluation results include: Adopting an adaptive path growth algorithm, constructing a path tree from the starting point of the cable as an initial node, and marking the initial node as a current active node; Identifying candidate extension nodes adjacent to the current active node, checking obstacles of connecting segments between the current active node and the candidate extension nodes according to the power grid environment constraint map, and calculating load adaptability in combination with the cable capacity plan; Scoring the candidate expansion nodes based on the load adaptability according to the multi-objective evaluation function, selecting the candidate expansion node with the highest comprehensive score and adding it to the path tree as a new current active node; Calculate the potential index of the expansion direction formed between each active node and prune the low-potential branches; Repeat the above expansion process until the current active node reaches the end node or reaches the preset maximum number of expansion steps, and the obtained complete path and the comprehensive score of the nodes in the complete path are used as the path evaluation result.

7. The method for optimizing power grid cable routing according to claim 1, characterized in that: The performing resilience evaluation and optimization on the candidate path set under the preset multiple simulation scenario conditions to generate a power grid cable wiring plan includes: Calculating a path resilience index corresponding to each candidate path in the candidate path set under the preset multiple simulation scenario conditions; Calculate a comprehensive resilience score for each candidate path based on the path resilience index and the occurrence probability and impact intensity of the preset multiple simulation scenario conditions; Identify weak path segments of candidate paths whose comprehensive resilience scores are lower than a preset score threshold, and design backup paths and improve cable specifications for the weak path segments; A path sensitivity analysis is performed on the candidate path set under the preset multiple simulation scenario conditions, the stability index of the candidate path is calculated by adjusting the scenario parameters, the candidate path is selected according to the stability index, and a power grid cable wiring plan is generated.

8. A power grid cable routing optimization device, characterized in that: The power grid cable wiring path optimization device comprises: The load analysis module is used to perform load analysis on the geographical distribution data and power consumption data of the power grid system, and generate load area types and weight data corresponding to each area type; a risk assessment module, configured to perform a fault propagation risk assessment on grid nodes in the power grid system according to the load area type and weight data, and generate a cable capacity plan based on the risk assessment results; A path generation module, configured to generate a set of candidate paths according to the load area type and weight data and the cable capacity plan through a multi-objective balancing mechanism and an adaptive path growth algorithm; The resilience optimization module is used to evaluate and optimize the resilience of the candidate path set under multiple preset simulation scenario conditions to generate a power grid cable wiring plan.

Images