Node weight-based power material vehicle-unmanned aerial vehicle cooperative scheduling method and system

Abstract

This application discloses a power supply vehicle-UAV collaborative scheduling method and system based on node weights, relating to the fields of transportation and logistics technology. The method includes: forming a basic dataset by collecting multi-source data such as power grid topology and emergency demand points; calculating the comprehensive weight of demand nodes by integrating power grid topology and emergency demand information; constructing and solving a collaborative scheduling optimization model by combining transport capacity and real-time environmental data; generating the optimal scheduling scheme; and sending control commands to the vehicle and UAV to execute the collaborative scheduling. By prioritizing high-priority nodes, reducing weighted delivery time, and controlling transportation costs, the method enhances the adaptability of the scheme to road condition fluctuations, improves the efficiency and reliability of emergency scheduling, and achieves joint optimized scheduling of vehicles and UAVs.

Description

Technical Field

[0001] This invention relates to the field of transportation and logistics technology, and in particular to a method and system for collaborative scheduling of power supply vehicles and drones based on node weights. Background Technology

[0002] In power system emergency repair scenarios, the efficiency of post-disaster material dispatch directly affects the speed of grid recovery and the control of disaster losses. When natural disasters such as typhoons, floods, and earthquakes cause widespread damage to the power grid, a large number of nodes simultaneously generate material relief requests. These nodes differ significantly in their importance in the power grid topology, the severity of their damage, and their social impact, requiring scientific prioritization to achieve precise delivery. Meanwhile, vehicles possess heavy-load, long-distance transportation capabilities but lack flexibility, while drones have flexible shuttle capabilities but limited range and payload. Collaborating between the two can create complementary advantages.

[0003] Existing methods for dispatching emergency power supplies mainly employ uniform weight allocation or prioritize nodes solely based on distance, failing to integrate grid topology characteristics and real-time disaster information. This results in critical grid nodes or severely affected areas not receiving priority protection. Regarding vehicle-machine collaboration, existing methods often employ a two-stage approach: first planning vehicle routes, then allocating drone tasks. This lack of a joint optimization mechanism for vehicle routes and drone task allocation makes it difficult to fully utilize the flexible mobility of drones for precise last-mile delivery. In terms of optimization, existing algorithms typically use fixed penalty coefficients and mutation probabilities, failing to adaptively adjust based on population evolution and the satisfaction of high-weight nodes, leading to slow convergence or getting trapped in local optima. Furthermore, existing dispatch schemes lack consideration for uncertainties during execution, failing to account for the impact of fluctuating road conditions on delivery time. The generated schemes may fail in real-world emergency environments due to sudden changes in road conditions.

[0004] Therefore, how to integrate the importance of power grid topology, real-time disaster level and population density to construct a comprehensive node weight, and how to achieve joint optimization and adaptive solution of vehicle-UAV collaborative scheduling based on this weight, has become an urgent problem to be solved. Summary of the Invention

[0005] The main purpose of this application is to provide a method and system for collaborative scheduling of power material vehicles and drones based on node weights, aiming to solve the technical problem of how to integrate multi-source data to quantify node priorities and achieve joint optimization of vehicle-drone collaborative scheduling.

[0006] To achieve the above objectives, this application proposes a node weight-based collaborative scheduling method for power material vehicles and unmanned aerial vehicles, including: Collect power grid topology data, emergency demand point information, transportation capacity resource data, and real-time environmental dynamic data to obtain a basic dataset; The comprehensive weight of each demand node is calculated based on the power grid topology data and emergency demand point information in the basic dataset. A collaborative scheduling optimization model is constructed and solved based on the comprehensive weight, the transportation capacity resource data, and the real-time environmental dynamic data to obtain the optimal scheduling scheme. Control commands are generated based on the optimal scheduling scheme and sent to the vehicle and the drone to enable the vehicle and the drone to perform collaborative scheduling tasks.

[0007] In one embodiment, the step of calculating the comprehensive weight of each demand node based on the power grid topology data and emergency demand point information in the basic dataset includes: Extract the betweenness centrality, disaster severity level, and associated population density of each demand node from the aforementioned basic dataset; The betweenness centrality is normalized to obtain a topological importance index; The severity level of the disaster is adjusted according to a preset disaster adjustment index to obtain a disaster urgency index; The social benefit index is obtained by normalizing the associated population density. The topological importance index, the disaster urgency index, and the social benefit index are weighted and summed according to a preset topological balance coefficient to obtain a comprehensive weight.

[0008] In one embodiment, the step of constructing a collaborative scheduling optimization model based on the comprehensive weights, the transportation capacity resource data, and the real-time environmental dynamic data includes: The total delivery time is calculated based on the comprehensive weight and the road traffic coefficient in the real-time environmental dynamic data. The total transportation cost is calculated based on the vehicle unit distance fuel cost, vehicle total travel distance, drone unit energy consumption electricity cost, and drone total energy consumption in the aforementioned transportation resource data. Based on the aforementioned transportation resource data, the following constraints are calculated: vehicle capacity constraint penalty value, UAV range constraint penalty value, UAV payload constraint penalty value, and vehicle-UAV coordination time constraint penalty value. The total constraint violation penalty value is obtained by summing the vehicle capacity constraint penalty value, the UAV range constraint penalty value, the UAV payload constraint penalty value, and the vehicle-UAV coordination time constraint penalty value. The total transportation cost, total delivery time, and total constraint violation penalty value are linearly weighted according to preset timeliness weight coefficient, preset economy weight coefficient, and preset penalty weight coefficient to obtain the comprehensive optimization objective function value; A collaborative scheduling optimization model is constructed based on the comprehensive optimization objective function value.

[0009] In one embodiment, the step of calculating the vehicle capacity constraint penalty value, the UAV range constraint penalty value, the UAV payload constraint penalty value, and the vehicle-UAV coordination time constraint penalty value based on the transportation resource data includes: Extract the maximum load capacity of the vehicle from the transportation resource data, determine the actual load weight of the vehicle, and calculate the first difference between the actual load weight and the maximum load capacity. When the first difference is greater than the preset capacity threshold, the vehicle capacity constraint penalty value is calculated based on the first difference and the preset capacity penalty coefficient. Extract the maximum range of the UAV from the transportation resource data, determine the actual flight distance of the UAV, and calculate a second difference between the actual flight distance and the maximum range; When the second difference is greater than the preset range threshold, the range constraint penalty value of the UAV is calculated based on the second difference and the preset range penalty coefficient. Extract the maximum payload of the UAV from the transportation resource data, determine the actual payload weight of the UAV, and calculate the third difference between the actual payload weight and the maximum payload. When the third difference is greater than the preset load threshold, the load constraint penalty value of the UAV is calculated based on the third difference and the preset load penalty coefficient. Extract the vehicle's stopping time at the pick-up point from the transportation resource data, determine the total mission time of the drone, and calculate the fourth difference between the total mission time and the stopping time. When the fourth difference is greater than the preset collaboration threshold, the collaboration time constraint penalty value between the vehicle and the drone is calculated based on the fourth difference and the preset collaboration penalty coefficient.

[0010] In one embodiment, the step of constructing and solving a collaborative scheduling optimization model based on the comprehensive weights, the transportation capacity resource data, and the real-time environmental dynamic data to obtain the optimal scheduling scheme includes: The selection probability of each demand node is calculated based on the comprehensive weight, spatial distance in the emergency demand point information, and road traffic coefficient, flight environment coefficient, and disaster severity level change rate in the real-time environmental dynamic data. Based on the selection probabilities, an initial scheduling scheme is constructed using a stepwise insertion method to generate an initial population; The estimated delivery time is calculated based on the path planning and task allocation of each individual in the initial population. The adaptive mutation probability is calculated based on the comprehensive weight and the estimated delivery time. The individuals in the initial population are then subjected to mutation operation based on the adaptive mutation probability to obtain the mutated population. The penalty coefficient is dynamically adjusted based on the proportion of feasible solutions in the population and the constraint violation of high-weight nodes, and the cooperative scheduling optimization model is updated to obtain the updated cooperative scheduling optimization model. The mutated population is iteratively solved according to the updated cooperative scheduling optimization model to obtain a candidate solution set. Based on the candidate solution set, feasible solutions that satisfy the constraints are selected, and the optimal solution is chosen from the feasible solutions as the optimal scheduling scheme.

[0011] In one embodiment, the step of constructing an initial scheduling scheme and generating an initial population using a stepwise insertion method based on the selection probability includes: Obtain the set of demand nodes that are not currently being served, and randomly select a node to be served from the set of demand nodes according to the selection probability; Based on the maximum load capacity of vehicles in the transportation resource data, determine whether the node to be served is suitable for direct service by a vehicle. If it is suitable for direct service by a vehicle, insert the node to be served into the current vehicle path. If it is not suitable for direct service by a vehicle, mark the node to be served as a candidate node for drones. Based on the maximum range and maximum payload of the drone in the transportation resource data, determine whether the candidate drone node can be served by the drone from the current docking point. If it can be served by the drone from the current docking point, assign the candidate drone node to the corresponding drone task sequence. Until all demand nodes are allocated, the current scheduling scheme is constructed based on the driving paths of all vehicles and the task allocation scheme of all UAVs, and the current scheduling scheme is encoded as a population individual; Until the preset population size is reached, all individuals in the population are combined into an initial population.

[0012] In one embodiment, the steps of calculating the estimated delivery time based on the path planning and task allocation of each individual in the initial population, calculating the adaptive mutation probability based on the comprehensive weight and the estimated delivery time, and performing mutation operations on the individuals in the initial population according to the adaptive mutation probability to obtain the mutated population include: Obtain the first comprehensive weight of the current demand node and the estimated delivery time calculated based on the path planning and task allocation of the current demand node in the initial population; and obtain the second comprehensive weight of all demand nodes and the maximum delivery time of all individuals in the initial population. The relative weight value is obtained by calculating the ratio of the maximum value of the first comprehensive weight to the maximum value of the second comprehensive weight, and the relative timeliness value is obtained by calculating the ratio of the estimated delivery time to the maximum delivery time. The relative weight value and the relative time value are multiplied to obtain the weighted time product, and the difference between the weighted time product and the variation threshold is calculated according to the preset variation threshold to obtain the variation trigger deviation; The mutation triggering deviation is transformed by a function mapping based on a preset steepness coefficient to obtain a probability adjustment coefficient; The adaptive mutation probability is calculated based on the preset minimum mutation probability, the preset maximum mutation probability, and the probability adjustment coefficient. Based on the adaptive mutation probability, the gene loci in the initial population related to the current demand node are mutated to obtain the mutated population.

[0013] In one embodiment, the step of dynamically adjusting the penalty coefficient based on the proportion of feasible solutions in the population and the constraint violation of high-weight nodes to update the cooperative scheduling optimization model and obtain the updated cooperative scheduling optimization model includes: Obtain the number of individuals violating the constraints and the population size in the current population, and calculate the ratio of the number of individuals to the population size to obtain the violation ratio; Obtain a set of high-weight nodes, and obtain the weights of nodes in the set of high-weight nodes affected by vehicle overloading and the total weight of the set of high-weight nodes, wherein the set of high-weight nodes consists of nodes whose overall weight is greater than the average weight of all demand nodes. The high-weight violation rate is obtained by calculating the ratio of the node weight to the total weight. The violation ratio is linearly adjusted according to a preset linear adjustment factor to obtain the adjusted violation ratio. The high-weight violation rate is adjusted exponentially according to a preset exponential adjustment factor to obtain the adjusted high-weight violation rate. The preset initial penalty coefficient is dynamically adjusted based on the adjusted violation ratio and the adjusted high-weight violation rate to obtain the dynamic penalty coefficient. The vehicle capacity constraint penalty coefficient, UAV range constraint penalty coefficient, UAV payload constraint penalty coefficient, and vehicle-UAV coordination time constraint penalty coefficient in the collaborative scheduling optimization model are updated based on the dynamic penalty coefficient to obtain the updated collaborative scheduling optimization model.

[0014] In one embodiment, the step of filtering feasible solutions that satisfy the constraints based on the candidate solution set and selecting the optimal solution from the feasible solutions as the optimal scheduling scheme includes: The total constraint violation penalty value for each candidate solution is calculated based on the candidate solution set. When the total constraint violation penalty value is less than the preset feasible threshold, the candidate solution is determined as a feasible solution, and all the feasible solutions are combined into a feasible solution set; For each feasible solution in the feasible solution set, the weighted total delivery time is calculated based on the comprehensive weight and the delivery time of each demand node in the feasible solution; The total transportation cost is calculated based on the fuel cost per unit distance of the vehicle, the total distance traveled by the vehicle, the unit energy cost of the drone, and the total energy consumption of the drone. A robustness penalty term is calculated based on the sensitivity of the delivery time of each demand node to fluctuations in the road traffic coefficient. The weighted total delivery time, the total transportation cost, and the robustness penalty term are weighted and summed according to the preset robustness weight coefficient to obtain a comprehensive robustness index. The feasible solution with the smallest comprehensive robustness index is selected from the feasible solution set as the optimal scheduling scheme.

[0015] Furthermore, to achieve the above objectives, this application also proposes a power material vehicle-UAV collaborative scheduling system based on node weights, wherein the power material vehicle-UAV collaborative scheduling system based on node weights includes: The data acquisition module is used to collect power grid topology data, emergency demand point information, transportation capacity resource data, and real-time environmental dynamic data to obtain a basic dataset. The weight calculation module is used to calculate the comprehensive weight of each demand node based on the power grid topology data and emergency demand point information in the basic dataset. The results module is used to construct and solve a collaborative scheduling optimization model based on the comprehensive weight, the transportation capacity resource data, and the real-time environmental dynamic data to obtain the optimal scheduling scheme. The sending module is used to generate control commands according to the optimal scheduling scheme and send the control commands to the vehicle and the drone so that the vehicle and the drone can perform collaborative scheduling tasks.

[0016] This application forms a basic dataset by collecting multi-source data such as power grid topology and emergency demand points. It then integrates power grid topology and emergency demand information to calculate the comprehensive weight of demand nodes. Finally, it combines transport capacity and real-time environmental data to construct and solve a collaborative scheduling optimization model. After generating the optimal scheduling scheme, it sends control commands to the vehicle-to-machine (V2M) system to execute the collaborative scheduling. By prioritizing high-priority nodes, reducing weighted delivery time, and controlling transportation costs, the application enhances the adaptability of the scheme to road condition fluctuations, improves the efficiency and reliability of emergency dispatching, and achieves joint optimized scheduling between vehicles and the V2M system. Attached Figure Description

[0017] To more clearly illustrate the technical solutions in the embodiments of this application or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, for those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0018] Figure 1 This is a flowchart illustrating the first embodiment of the power material vehicle-UAV collaborative scheduling method based on node weights in this application; Figure 2 This is a flowchart illustrating the second embodiment of the power material vehicle-UAV collaborative scheduling method based on node weights in this application; Figure 3 This is a schematic diagram of the module structure of the power material vehicle-UAV collaborative scheduling system based on node weights in this application; Figure 4 This is a schematic diagram of the equipment structure of the hardware operating environment involved in the power material vehicle-UAV collaborative scheduling method based on node weight in the embodiments of this application.

[0019] The purpose, features, and advantages of this application will be further explained in conjunction with the embodiments and with reference to the accompanying drawings. Detailed Implementation

[0020] It should be understood that the specific embodiments described herein are merely illustrative of the technical solutions of this application and are not intended to limit this application.

[0021] To better understand the technical solution of this application, a detailed description will be provided below in conjunction with the accompanying drawings and specific implementation methods.

[0022] A basic dataset is formed by collecting multi-source data such as power grid topology and emergency demand points. The comprehensive weight of demand nodes is calculated by integrating power grid topology and emergency demand information. Then, a collaborative scheduling optimization model is constructed and solved by combining capacity and real-time environmental data. After generating the optimal scheduling scheme, control commands are sent to the vehicle-to-machine interface to execute the collaborative scheduling. By prioritizing high-priority nodes, reducing weighted delivery time, and controlling transportation costs, the scheme enhances its adaptability to road condition fluctuations, improves the efficiency and reliability of emergency scheduling, and achieves joint optimized scheduling between vehicles and the system.

[0023] Based on the above, this application also provides a method for collaborative scheduling of power material vehicles and drones based on node weights, referring to... Figure 1 , Figure 1 This is a flowchart illustrating the first embodiment of the power material vehicle-UAV collaborative scheduling method based on node weights in this application.

[0024] In this embodiment, the power material vehicle-UAV collaborative scheduling method based on node weight includes steps S10~S40: Step S10: Collect power grid topology data, emergency demand point information, transportation capacity resource data, and real-time environmental dynamic data to obtain the basic dataset.

[0025] Specifically, firstly, the geographic coordinates, connectivity relationships, and network structure parameters of all power grid nodes are obtained from the power grid geographic information system as power grid topology data. This is done to calculate topological importance indicators such as betweenness centrality of each node. Secondly, the specific location, disaster severity rating, population density of the associated area, and quantity and type of materials required for each demand point are collected through disaster on-site reporting and remote sensing image analysis as emergency demand point information. This is done to quantify the urgency and social impact of each node in an emergency scenario. Then, material inventory information, the number and maximum load capacity of available vehicles, and the number, maximum range, and maximum load of available drones are collected from material reserve points as transportation resource data. This is done to clarify the schedulable resource boundaries and capacity constraints. Finally, real-time road traffic coefficients, disaster severity rating change rates, and drone flight environment coefficients are obtained from the traffic information platform and meteorological service system as real-time environmental dynamic data. This is done to ensure that the scheduling scheme can adapt to road condition fluctuations and weather changes, ensuring the feasibility of the scheme in the actual execution environment. The above four types of data are integrated to obtain a basic dataset, laying the data foundation for subsequent node weight calculation and collaborative scheduling optimization.

[0026] Step S20: Calculate the comprehensive weight of each demand node based on the power grid topology data and emergency demand point information in the basic dataset.

[0027] It should be noted that step S20 includes: extracting the betweenness centrality, disaster severity level, and associated population density of each demand node from the basic dataset; normalizing the betweenness centrality to obtain the topological importance index; adjusting the disaster severity level according to a preset disaster adjustment index to obtain the disaster urgency index; normalizing the associated population density to obtain the social benefit index; and weighting and summing the topological importance index, disaster urgency index, and social benefit index according to a preset topological balance coefficient to obtain the comprehensive weight.

[0028] It's important to understand that betweenness centrality is an indicator that measures the importance of a power grid node in the network topology. It characterizes the importance of a node as the shortest path bridge in the power grid topology, and its value is directly calculated from the power grid topology structure. It is a core quantitative indicator of a node's topological characteristics. Disaster severity level is a quantitative score of the severity of the disaster in the area where the demand node is located. It is usually a discrete integer; a higher level indicates a more severe disaster in the node's area. It is obtained from on-site assessments or remote sensing data and is a key indicator reflecting the urgency of a node's emergency response. Associated population density is the population distribution density within the area associated with or influenced by the demand node, measured in people per square kilometer. It reflects the scale of social benefits that the node can generate by carrying out material relief and is a core indicator for measuring the node's social impact characteristics. The preset topology balance coefficient is a hyperparameter used to balance the proportion of topology importance indicators with disaster urgency and social benefit indicators in the overall weighting. Its value ranges from 0 to 1 and includes sub-coefficients for balancing disaster urgency and social benefit indicators, used to allocate the weight proportions of different characteristic indicators.

[0029] Specifically, firstly, the betweenness centrality, disaster severity score, and associated population density of each demand node are retrieved from the basic dataset as raw feature data. Secondly, the betweenness centrality values ​​of all nodes are divided by the maximum value to obtain a normalized value, which is then mapped to a preset interval using a hyperbolic tangent function to obtain a topological importance index. This is done to eliminate dimensional differences and prevent the topological index value from being too large and excessively dominating the weight calculation. Next, the disaster severity score is exponentially calculated, and a preset disaster adjustment index is used as the index to obtain a disaster urgency index. This is done to highlight the urgency of high-level disaster points through a non-linear amplification mechanism. Subsequently, the associated population density is divided by the maximum value and normalized to obtain a social benefit index. This is done to quantify the scale of social benefits generated by serving this node. Finally, the topological importance index is multiplied by a preset topological balance coefficient, and the disaster urgency index and social benefit index are weighted and multiplied by the remaining weight coefficient. The sum of the two parts is then obtained to obtain the comprehensive weight. The specific formula is as follows: in Indicates the first The comprehensive weight of each demand node represents its priority order in the global emergency dispatch. The larger the value, the more priority the node should receive in the dispatch scheme. Indicates the first The betweenness centrality of a demand node characterizes the importance of that node as a shortest path bridge in the power grid topology; its value is calculated from the power grid topology. This indicates the betweenness centrality among all demand nodes. The maximum value; The adjustment hyperparameter represents the balance between topological importance and real-time disaster situation and social impact, and its value ranges from [0,1]. Indicates the first Disaster severity rating of each demand node of Power of 1 Used to quantify the real-time disaster urgency of the area where the node is located, it is usually obtained from on-site assessment or remote sensing data. Example values ​​are provided. , The nonlinear adjustment hyperparameter representing the disaster severity level is a real number greater than zero. The weighting of high-level disaster points has an amplifying effect; This represents the adjustment hyperparameter that balances the disaster severity factor and population density factor in real-time disaster information, and its value range is [value range missing]. ; Indicates the first The population density of the area associated with each demand node is used to measure the scale of social benefits that can be generated by serving that node, in units of... , This represents the population density among all demand nodes. The maximum value; This represents the index of the demand node, which is a positive integer and its value ranges from 1 to 2. This is used to uniquely identify each node that needs material assistance. This represents the total number of required nodes, determined by the total number of points awaiting rescue in the emergency scenario; This represents the hyperbolic tangent function. This is done to balance the importance of the power grid topology with the real-time social impact of disasters, ensuring that dispatching schemes prioritize critical nodes while also considering severely affected and densely populated areas.

[0030] Step S30: Construct and solve a collaborative scheduling optimization model based on comprehensive weights, transportation capacity resource data, and real-time environmental dynamic data to obtain the optimal scheduling scheme.

[0031] It should be noted that step S30 includes: First, calculating the total delivery time based on the comprehensive weight and the road traffic coefficient in the real-time environmental dynamic data. Specifically, for each demand node, the actual travel distance from the material reserve point to the node is calculated based on the assigned vehicle or drone and the road segments traversed in the route planning; second, the actual travel time of the node is calculated based on the road traffic coefficient obtained from the real-time environmental dynamic data (this coefficient reflects the smoothness of the road segment, and the closer the value is to 1, the smoother the road) and the preset average travel speed of the vehicle (travel distance divided by the product of speed and road traffic coefficient). This is done to convert real-time road condition fluctuations into quantifiable travel time delays; then, the actual travel time of the node is multiplied by its comprehensive weight to obtain the weighted delivery time of the node. This is done to ensure that the time cost of high-priority nodes accounts for a larger proportion in the objective function; finally, the weighted delivery times of all demand nodes are summed to obtain the total delivery time as the timeliness objective function value of the collaborative scheduling optimization model. This is done to drive the optimization algorithm to prioritize the timely delivery of materials to high-weight nodes during the search process.

[0032] Next, based on the vehicle unit distance fuel cost, total vehicle travel distance, drone unit energy cost, and total drone energy consumption data, the total transportation cost is calculated. Specifically, firstly, for each vehicle participating in the dispatch, its unit distance fuel cost is multiplied by its total travel distance to obtain the vehicle's fuel cost. This is done to quantify the economic consumption of long-distance vehicle transportation. Secondly, for each drone participating in the dispatch, its unit energy cost is multiplied by its total energy consumption to obtain the drone's electricity cost. This is done to quantify the economic consumption of drone last-mile delivery. Then, the fuel costs of all vehicles are summed to obtain the total vehicle cost, and the electricity costs of all drones are summed to obtain the total drone cost. Finally, the total vehicle cost and the total drone cost are summed to obtain the total transportation cost. The specific formula is as follows: in This represents the total transportation cost. This represents the vehicle index, a positive integer ranging from {1, 2, ..., V}, used to uniquely identify each vehicle participating in the scheduling. The total number of vehicles participating in the dispatch is determined by the emergency supply reserve points and available transport capacity. This indicates the fuel cost per unit distance for a vehicle. Indicates the total distance traveled by the vehicle; This represents the drone index, which is a positive integer with a value range of 1. Used to uniquely identify each drone participating in the scheduling. The total number of drones participating in the dispatch is determined by the emergency material reserve points and available transportation capacity. This indicates the unit energy consumption cost of a drone. This indicates the total energy consumption of the drone.

[0033] Next, based on the transportation resource data, calculations are performed to obtain vehicle capacity constraint penalty values, drone range constraint penalty values, drone payload constraint penalty values, and vehicle-drone coordination time constraint penalty values. Further, the maximum payload capacity of the vehicle is extracted from the transportation resource data, and the actual load weight of the vehicle is determined. The first difference between the actual load weight and the maximum payload capacity is calculated. When the first difference is greater than a preset capacity threshold, the vehicle capacity constraint penalty value is calculated based on the first difference and a preset capacity penalty coefficient. The specific formula is as follows: in, Indicates the first The penalty value for a vehicle violating capacity constraints; The penalty coefficient for capacity constraints is a positive constant used to control the strength of the penalty. The larger the coefficient, the more severe the penalty for violating the capacity constraint. Indicates the first The actual total load weight of the vehicle is a decision variable obtained after solving the model, and the unit is kilograms; Indicates the first The maximum load capacity of a vehicle is a known inherent parameter of the vehicle, and the unit is kilograms. The function is used to ensure that only when the actual load exceeds the maximum capacity, i.e. Only then will a positive penalty value be generated.

[0034] Extract the maximum range of the UAV from the capacity resource data, determine the actual flight distance of the UAV, and calculate the second difference between the actual flight distance and the maximum range. When the second difference is greater than a preset range threshold, calculate the UAV range constraint penalty value based on the second difference and a preset range penalty coefficient. The specific formula is as follows: in, Indicates the first The penalty value for a drone violating range constraints; This represents the penalty coefficient for the range constraint. It is a positive constant used to control the strength of this penalty. An example value is shown below. ; Indicates the first The actual total flight distance of the drone is a decision variable obtained after solving the model, and the unit is kilometers; Indicates the first The maximum range of the drone is given by the known inherent parameters of the drone, in kilometers.

[0035] Extract the maximum payload of the drone from the capacity resource data, determine the actual payload weight of the drone, and calculate the third difference between the actual payload weight and the maximum payload. When the third difference is greater than a preset payload threshold, calculate the drone payload constraint penalty value based on the third difference and a preset payload penalty coefficient. The specific formula is as follows: in, Indicates the first The penalty value for a drone violating load constraints; The penalty coefficient representing the load constraint is a positive constant used to control the intensity of this penalty. Indicates the first The actual payload weight of the drone is a decision variable obtained after solving the model, and the unit is kilograms; Indicates the first The maximum payload of the drone is given by the known inherent parameters of the drone, and the unit is kilograms.

[0036] The system extracts the vehicle's dwell time at the docking point from the transportation resource data and determines the total mission time of the drone. It then calculates the fourth difference between the total mission time and the dwell time. The docking point is the connection point for the coordinated dispatch of emergency power supplies between the vehicle and the drone. It is the designated docking point where the drone retrieves supplies from the vehicle, completes delivery, and returns to the vehicle. It is also the operation point where the vehicle provides supplies to the drone and completes the drone's recovery. When the fourth difference is greater than a preset coordination threshold, a time constraint penalty value for vehicle-drone coordination is calculated based on the fourth difference and a preset coordination penalty coefficient. The specific formula is as follows: in, Indicates the first The penalty value for a drone violating the collaborative time constraint; The penalty coefficient representing the cooperative time constraint is a positive constant used to control the strength of this penalty. Indicates the first A drone from the docking point Take off, serve all assigned nodes and return to the pick-up point. The total time required is the decision variable obtained after solving the model, in hours; Indicates the first The first drone relies on The vehicle at the pick-up point The stop duration is a decision variable obtained after solving the model, and the unit is hours; The function is used to ensure that the drone mission's total time exceeds the vehicle's parking time only if the total time is greater than the vehicle's parking time. Only then will a positive penalty value be generated.

[0037] Then, the total constraint violation penalty value is obtained by summing the vehicle capacity constraint penalty value, the UAV range constraint penalty value, the UAV payload constraint penalty value, and the vehicle-UAV coordination time constraint penalty value. The specific formula is as follows: in, Indicates the first Vehicle capacity constraint penalty value; Indicates the first The drone's range constraint penalty value; Indicates the first The drone load constraint penalty value for the drone; Indicates the first The penalty value for the time constraint of vehicle-drone collaboration with drones.

[0038] Then, based on the preset timeliness weight coefficient, preset economy weight coefficient, and preset penalty weight coefficient, the total transportation cost, total delivery time, and total constraint violation penalty value are linearly weighted to obtain the comprehensive optimization objective function value. The specific formula is as follows: in This represents the overall optimization objective function value, which is the overall goal of the model optimization. The smaller the value, the better the overall performance of the obtained scheduling scheme in terms of timeliness, economy, and feasibility. This means minimizing the value of the integrated objective function, i.e., seeking to make... The scheduling scheme with the minimum value; Indicates total delivery time The weight coefficients in the overall objective function are real numbers greater than or equal to zero, used to control the relative importance of timeliness objectives in the overall optimization; The total delivery time represents the core timeliness objective for optimization. A smaller value indicates more timely delivery of materials to high-priority nodes. The calculation method is expressed as follows: ; Indicates total transportation cost The weight coefficients in the overall objective function are real numbers greater than or equal to zero, used to control the relative importance of economic objectives in the overall optimization. This represents the total transportation cost; Indicates penalties for violating constraints The weight coefficients in the overall objective function are real numbers greater than or equal to zero. They are used to control the penalty for violations of constraints and encourage the algorithm to prioritize searching for feasible solutions. This indicates a penalty for violating the general constraint; Indicates that the supplies have been delivered to the first The actual time of each demand node refers to the time length, measured in hours, from the start of unified scheduling until the demand for resources at that node is met. Finally, a collaborative scheduling optimization model is constructed based on the comprehensive optimization objective function value.

[0039] Step S40: Generate control commands based on the optimal scheduling scheme and send the control commands to the vehicle and the drone so that the vehicle and the drone can perform collaborative scheduling tasks.

[0040] It should be noted that, firstly, the optimal scheduling scheme is parsed into an executable sequence of control commands. Specifically, this includes generating driving commands for each vehicle containing the route, stop locations along the way, and corresponding time windows, and generating task commands for each drone containing takeoff and pick-up points, a list of delivery points, and the type and quantity of supplies. This is to transform the abstract mathematical optimization results into a sequence of physical actions that vehicles and drones can directly execute. Secondly, vehicle control commands and drone control commands are simultaneously sent to the vehicle terminal and drone flight control system, triggering vehicles and drones to depart synchronously from the supply depot. This is to ensure the time consistency of vehicle-drone collaborative operations. Then, the vehicles proceed according to the driving instructions... The instructions are to transport supplies to a pre-designated rendezvous point and wait, while simultaneously releasing the onboard drones. The drones, following mission instructions, take off from the rendezvous point, fly to the designated demand point, deliver the supplies, and return to the same rendezvous point. The vehicles then retrieve the supplies and proceed to the next rendezvous point or return to the storage point. This is done to fully leverage the combined advantages of long-distance vehicle transport and precise last-mile delivery by drones. Finally, a remote monitoring system tracks the location, status, and remaining supplies of vehicles and drones in real time. When road congestion, sudden weather changes, or new urgent needs are detected, a dynamic adjustment mechanism is activated to recalculate local routes or task allocations and update control instructions. This ensures that supplies at critical nodes can still be delivered in a timely manner even in complex emergency environments.

[0041] This embodiment uses multi-source data, including power grid topology and emergency demand points, to form a basic dataset. It then integrates power grid topology and emergency demand information to calculate the comprehensive weight of demand nodes. Combined with capacity and real-time environmental data, it constructs and solves a collaborative scheduling optimization model. After generating the optimal scheduling scheme, it sends control commands to the vehicle-to-machine (V2M) system to execute the collaborative scheduling. By prioritizing high-priority nodes, reducing weighted delivery time, and controlling transportation costs, it enhances the scheme's adaptability to road condition fluctuations, improves emergency dispatch efficiency and reliability, and achieves joint optimized scheduling between vehicles and the V2M system.

[0042] Based on the first embodiment of this application, in the second embodiment of this application, the content that is the same as or similar to that in Embodiment 1 above can be referred to the above description, and will not be repeated hereafter. Based on this, please refer to... Figure 2 The method for collaborative scheduling of power supply vehicles and drones based on node weights, step S30, further includes steps S201 to S206: Step S201: Calculate the selection probability of each demand node based on the comprehensive weight, spatial distance in the emergency demand point information, and road traffic coefficient, flight environment coefficient, and disaster severity level change rate in the real-time environmental dynamic data.

[0043] Specifically, for each demand node that is not currently being served, the selection probability is obtained based on a comprehensive weight, spatial distance in the emergency demand point information, and road traffic coefficient, flight environment coefficient, and disaster severity level change rate in real-time environmental dynamic data. The specific formula is as follows: In the formula, This indicates that the first step is selected in the current construction step. The probability that a demand node will become the next service node; the higher the probability value, the greater the likelihood that the node will be selected. Indicates the first The comprehensive weight of each demand node is used to quantify the priority of that node in emergency dispatch; the larger the value, the higher the priority. The adjustment index, representing the overall weight of a node, is an adjustable hyperparameter used to control the strength of the influence of node priority on the selection probability. Indicates the first The spatial distance from a demand node to a material storage point is calculated from the coordinates of the node and the storage point, in kilometers. Its reciprocal is used to characterize the attraction of the distance; the closer the distance, the larger the value. The adjustment index representing node distance is an adjustable hyperparameter used to control the degree of influence of distance factors on selection probability. Indicates the first The average road traffic coefficient of the area where each demand node is located is obtained through real-time traffic information, and its value ranges from [value range missing]. The closer the value is to 1, the smoother the road is; The adjustment index, representing the road traffic coefficient, is an adjustable hyperparameter used to control the degree of influence of this coefficient on the selection probability. Indicates the first The drone flight environment coefficient for the area where each demand node is located is obtained through real-time meteorological information, and its value range is [value range missing]. The closer the value is to 1, the more suitable the flight environment is; The adjustment index, representing the drone flight environment coefficient, is an adjustable hyperparameter used to control the degree of influence of this coefficient on the selection probability. Indicates the first The rate of change of the disaster severity level of each demand node over time. This rate of change is calculated from real-time disaster monitoring data. The value can be positive or negative, and the unit is the level per hour. A positive value indicates that the disaster is worsening. The adjustment coefficient, representing the rate of change of the severity level of the disaster, is an adjustable hyperparameter used to control the degree of influence of the rate of change on the selection probability. Represents the natural constant; This represents the set of all demand nodes that have not yet been served by any vehicle or drone. This is done so that the sum of the selection probabilities of all nodes is 1, forming a valid probability distribution. This allows for random node selection based on multi-factor comprehensive evaluation in the stepwise insertion method, prioritizing nodes with high weight, proximity, good road conditions, favorable weather, and exacerbated disasters to enter the initial scheduling scheme.

[0044] Step S202: Construct an initial scheduling scheme using the stepwise insertion method based on the selection probability to generate an initial population.

[0045] It should be noted that step S202 includes: obtaining a set of demand nodes that are not currently being served, and randomly selecting nodes to be served from the demand node set according to the selection probability; determining whether the nodes to be served are suitable for direct service by vehicles based on the maximum load capacity of vehicles in the transportation resource data; if suitable for direct service by vehicles, inserting the nodes to be served into the current vehicle path, or if unsuitable for direct service by vehicles, marking the nodes to be served as drone candidate nodes; determining whether the drone candidate nodes can be served by drones from the current connection point based on the maximum range and maximum load of drones in the transportation resource data; if they can be served by drones from the current connection point, assigning the drone candidate nodes to the corresponding drone task sequence; until all demand nodes are assigned, constructing the current scheduling scheme based on the driving paths of all vehicles and the task allocation scheme of all drones, and encoding the current scheduling scheme into population individuals; until the preset population size is reached, forming an initial population from all population individuals.

[0046] Specifically, firstly, based on a set of all demand nodes not yet served by any vehicle or drone, a node to be served is randomly selected from this set according to selection probability. This is done to ensure random diversity while giving high-priority nodes a greater chance of being selected first. Secondly, based on transportation resource data, it is determined whether the node is suitable for direct vehicle service. If the vehicle's remaining load capacity is sufficient to accommodate the node's material needs, it is inserted at the end of the current vehicle's route. If the vehicle's capacity is insufficient or the node's location is outside the vehicle's direct range, it is marked as a drone candidate node. This is done to ensure that vehicle capacity constraints are not violated and to reasonably distinguish the service boundaries between vehicles and drones. Then, for drone candidate nodes, it is determined whether their material demand does not exceed the drone's maximum load capacity and whether the node... If the round-trip distance to the current docking point does not exceed the drone's maximum range, the node is assigned to the drone mission sequence starting from the current docking point. If the range or payload is exceeded, the process returns to select another node or starts a new path. This is done to ensure that the drone's physical constraints are met and to achieve accurate last-mile delivery. Finally, the above node selection and allocation steps are repeated until all required nodes are assigned. The currently formed set of vehicle paths and drone mission allocation scheme are encoded into a population, and the entire construction process is repeated until the preset population size is reached. All individuals form the initial population. This is done to generate a set of high-quality initial solutions that satisfy the constraints and have genetic diversity, providing an excellent search starting point for subsequent evolutionary algorithm iterations.

[0047] Step S203: Calculate the estimated delivery time based on the path planning and task allocation of each individual in the initial population, calculate the adaptive mutation probability based on the comprehensive weight and the estimated delivery time, and perform mutation operation on the individuals in the initial population based on the adaptive mutation probability to obtain the mutated population.

[0048] It should be noted that step S203 includes: obtaining the first comprehensive weight of the current demand node and the estimated delivery time calculated based on the path planning and task allocation of the current demand node in the initial population; obtaining the second comprehensive weight of all demand nodes and the maximum delivery time of all individuals in the initial population; calculating the ratio of the maximum value of the first comprehensive weight to the second comprehensive weight to obtain the relative weight value; calculating the ratio of the estimated delivery time to the maximum delivery time to obtain the relative timeliness value; multiplying the relative weight value and the relative timeliness value to obtain the weight-timeliness product; calculating the difference between the weight-timeliness product and the mutation threshold based on the preset mutation threshold to obtain the mutation triggering deviation; performing a function mapping transformation on the mutation triggering deviation based on the preset steepness coefficient to obtain the probability adjustment coefficient; calculating the adaptive mutation probability based on the preset minimum mutation probability, the preset maximum mutation probability, and the probability adjustment coefficient; and performing mutation operations on the gene loci related to the current demand node in the initial population based on the adaptive mutation probability to obtain the mutated population.

[0049] Specifically, firstly, for the current pending demand node, its comprehensive weight is read from the basic dataset as the first comprehensive weight. Then, the estimated completion time from the scheduling starting point to the node is calculated from the path planning and task allocation schemes of the current population individuals for that node, serving as the estimated delivery time. Simultaneously, all demand nodes are traversed to obtain the largest comprehensive weight as the second comprehensive weight, and all individuals in the current population are traversed to obtain the largest estimated delivery time as the maximum delivery time. This is done to establish the relative relationship between local node features and global extrema. Secondly, the first comprehensive weight is divided by the second comprehensive weight for normalization to obtain a relative weight value, and the estimated delivery time is divided by the maximum delivery time for normalization to obtain a relative timeliness value. This is done to eliminate dimensional differences, bringing weights and time to the same numerical scale for easier coupled calculations. Then, the relative weight value is multiplied by the relative time-dependent value to obtain the weight-time product. This product is subtracted from a preset mutation threshold to obtain the mutation triggering bias. This is done to quantify the urgency of the node's current optimization state; a large bias indicates urgent optimization needs when the weight is high and delivery is late. Next, the mutation triggering bias is multiplied by a preset steepness coefficient and then input into an S-shaped function for mapping transformation to obtain the probability adjustment coefficient. This is done to convert linear bias into a smooth probability adjustment factor, avoiding probability abrupt changes that could lead to search instability. Finally, the probability adjustment coefficient is scaled against a preset maximum and minimum mutation probability interval to obtain the adaptive mutation probability, as shown in the following formula: in Indicates the first The adaptive mutation probability of the gene locus corresponding to each demand node is used to control the possibility of random adjustments to the decision related to that node. This indicates the preset minimum mutation probability; This indicates the preset maximum mutation probability. ; Represents the natural exponential function; The steepness coefficient of the Sigmoid function is an adjustable hyperparameter used to control the rate at which the mutation probability changes with the independent variable. This represents the maximum value of the overall weight across all demand nodes, used to normalize the weight factors and make the ratio... The range of values ​​is ; This indicates that in the scheduling scheme represented by the current individual, the [number]th [individual / individual]... The delivery time of materials to each demand node is calculated from the path and task allocation in the current solution, in hours; This represents the maximum resource delivery time for a single individual among all individuals in the current population. It is used to normalize delivery times, making the ratio [value missing]. The range of values ​​is ; The threshold offset parameter of the Sigmoid function is an adjustable hyperparameter used to adjust the center point of the probability change curve. This is done to give high-weight and high-latency nodes a higher mutation probability to explore better solutions. Finally, a random number is generated and compared with this adaptive mutation probability. When the random number is less than the probability value, the gene locus of that node in the current individual is mutated. This process is repeated to traverse all nodes and individuals to obtain the mutated population. This is done to prioritize perturbing the configuration of those critical but high-latency nodes, accelerating convergence towards higher-priority nodes while maintaining population diversity.

[0050] Step S204: Dynamically adjust the penalty coefficient based on the proportion of feasible solutions in the population and the constraint violation of high-weight nodes, update the collaborative scheduling optimization model, and obtain the updated collaborative scheduling optimization model.

[0051] It should be noted that step S204 includes: obtaining the number of individuals violating constraints and the population size in the current population, and calculating the ratio of the number of individuals to the population size to obtain the violation ratio; obtaining the set of high-weight nodes, and obtaining the weights of nodes affected by vehicle overloading and the total weight of the high-weight node set, wherein the high-weight node set consists of nodes whose comprehensive weight is greater than the average weight of all demand nodes; calculating the ratio of node weights to total weight to obtain the high-weight violation rate; linearly adjusting the violation ratio according to a preset linear adjustment factor to obtain the adjusted violation ratio; exponentially adjusting the high-weight violation rate according to a preset exponential adjustment factor to obtain the adjusted high-weight violation rate; dynamically adjusting the preset initial penalty coefficient according to the adjusted violation ratio and the adjusted high-weight violation rate to obtain the dynamic penalty coefficient; updating the vehicle capacity constraint penalty coefficient, UAV range constraint penalty coefficient, UAV payload constraint penalty coefficient, and vehicle-UAV coordination time constraint penalty coefficient in the collaborative scheduling optimization model according to the dynamic penalty coefficient to obtain the updated collaborative scheduling optimization model.

[0052] Specifically, in this embodiment, taking the penalty coefficient for vehicle capacity constraints as an example, firstly, the number of individuals in the current population whose actual load weight exceeds the maximum load capacity is counted as the number of individuals violating the constraint, and the total number of individuals in the current population is obtained as the population size. The number of violating individuals is divided by the population size to obtain the violation ratio. This is done to quantify the prevalence of infeasible solutions in the current population to determine the overall constraint pressure. Secondly, the average value of the comprehensive weight of all demand nodes is calculated as the weight threshold. Nodes with a comprehensive weight greater than this threshold are assigned to a high-weight node set, and the sum of the comprehensive weights of nodes in this set affected by the overloading of their respective vehicles is counted as the node weight sum. At the same time, the high-weight node set is also counted. The sum of the comprehensive weights of all nodes is taken as the total weight. Dividing the sum of the node weights by the total weight yields the high-weight violation rate. This is done to focus on the constraint satisfaction of key nodes rather than just the overall number of violations. Then, the violation ratio is multiplied by a preset linear adjustment factor to amplify it, resulting in an adjusted violation ratio. The high-weight violation rate is then multiplied by a preset exponential adjustment factor and subjected to exponential operation to obtain the adjusted high-weight violation rate. This is done so that violations at the population level affect the penalty intensity linearly, while violations at the high-weight level affect it non-linearly. Finally, the adjusted violation ratio is multiplied by the adjusted high-weight violation rate, then 1 is added, and finally multiplied by a preset initial penalty coefficient to obtain the dynamic penalty coefficient. The specific formula is as follows: in, Indicates the first The larger the value of the vehicle capacity constraint penalty coefficient used in the next iteration, the more severe the penalty for violating the capacity constraint. This represents the preset initial vehicle capacity constraint penalty coefficient, which serves as the benchmark value for dynamic adjustment. This represents a linear adjustment factor based on the proportion of individuals violating the previous generation. It is an adjustable hyperparameter used to control the strength of this part's influence on the penalty coefficient. Indicates the first The number of individuals that violate the vehicle capacity constraint in the next iteration of the population reflects the prevalence of the violation of the constraint in the previous generation of the population. The population size, i.e. the total number of individuals in the population, is a pre-defined constant. This represents the exponential adjustment factor based on the overload situation of high-weight nodes in the current generation. It is an adjustable hyperparameter used to control the intensity of this part's influence on the penalty coefficient. This represents a set of nodes with high weight requirements. This set typically consists of nodes with weights greater than the average weight of all nodes, and is used to focus on the constraint satisfaction of key nodes. This indicates an indicator function; the function value is 1 when the logical expression within the parentheses is true, and 0 otherwise. This indicates that in the scheduling scheme represented by a certain individual, the one responsible for serving the [number]th [unit / item] is [position / role]. The demand node's first The vehicle's actual total load weight is calculated by summing up the vehicle's delivery tasks, and the unit is kilograms; Indicates the first The maximum load capacity of a vehicle is a known inherent parameter of the vehicle, and the unit is kilograms. This represents the summation of the weights of all nodes in the high-weight node set that are affected by the overloading of their respective vehicles, used to quantify the overall degree to which high-weight nodes are affected by overloading. This represents the sum of the weights of all nodes in the high-weight node set, used to normalize the aforementioned affected weights. This is done to automatically increase the penalty intensity when widespread violations are severe or critical nodes are damaged, forcing the algorithm to converge towards the feasible region. Finally, dynamic penalty coefficients are assigned to the vehicle capacity constraint penalty coefficient, the UAV range constraint penalty coefficient, the UAV payload constraint penalty coefficient, and the vehicle-UAV coordination time constraint penalty coefficient, respectively, replacing the fixed penalty coefficients in the original model to obtain the updated collaborative scheduling optimization model. This is done to make the constraint penalty intensity adaptively adjust with the evolutionary state, avoiding premature convergence to the infeasible region or excessive penalty leading to search stagnation.

[0053] Step S205: Iteratively solve the mutated population according to the updated collaborative scheduling optimization model to obtain a candidate solution set.

[0054] Specifically, firstly, the mutated population is used as the current generation population. For each individual, the comprehensive optimization objective function value is recalculated as the fitness value based on the dynamic penalty coefficient in the updated cooperative scheduling optimization model. This is done to accurately assess the quality of the current solution using the adjusted constraint penalty strength. Secondly, tournament selection or roulette wheel selection operations are performed on the current generation population based on the fitness values. Individuals with better fitness values ​​are retained as parent individuals. Partial mapping crossover or sequential crossover operations are performed on the parent individuals to generate offspring individuals. Then, local mutation operations are performed on the offspring individuals based on the updated model parameters. This is done to maintain the fitness of the offspring while retaining the best individual fitness. While generating good genes, new solutions are generated and population diversity is maintained. Then, it is determined whether the preset termination condition is met, namely, reaching the maximum number of iterations or the change in the best fitness value over multiple generations is less than the preset convergence threshold. If the condition is met, all individuals in the current population are used as candidate solutions. If the condition is not met, the newly generated offspring individuals are used as the current generation population and the fitness value is recalculated. This is done to continuously optimize the quality of the solution through multiple generations of evolution until convergence. Finally, all individuals in the final population that meets the termination condition constitute the candidate solution set. This is done to obtain a set of candidate scheduling schemes that have been fully evolved and optimized for subsequent selection of the optimal solution.

[0055] Step S206: Select feasible solutions that meet the constraints based on the candidate solution set, and select the optimal solution from the feasible solutions as the optimal scheduling scheme.

[0056] It should be noted that step S206 includes: calculating the total constraint violation penalty value for each candidate solution based on the candidate solution set; when the total constraint violation penalty value is less than a preset feasible threshold, the candidate solution is determined as a feasible solution, and all feasible solutions are combined into a feasible solution set; for each feasible solution in the feasible solution set, the weighted total delivery time is calculated based on the comprehensive weight and the delivery time of each demand node in the feasible solution; the total transportation cost is calculated based on the fuel cost per unit distance of the vehicle, the total driving distance of the vehicle, the power cost per unit energy consumption of the drone, and the total energy consumption of the drone; the robustness penalty term is calculated based on the sensitivity of the delivery time of each demand node to the fluctuation of the road traffic coefficient; the weighted total delivery time, the total transportation cost, and the robustness penalty term are weighted and summed according to the preset robustness weight coefficient to obtain the comprehensive robustness index; and the feasible solution with the smallest comprehensive robustness index is selected from the feasible solution set as the optimal scheduling scheme.

[0057] Specifically, firstly, each candidate solution in the candidate solution set is traversed, and its vehicle capacity constraint penalty value, UAV range constraint penalty value, UAV payload constraint penalty value, and vehicle-UAV coordination time constraint penalty value are summed to obtain the total constraint violation penalty value. This is done to quantify the degree to which each candidate solution violates the physical constraints. Secondly, the total constraint violation penalty value is compared with a preset feasible threshold. When the value is less than the preset feasible threshold, the candidate solution is marked as a feasible solution. All feasible solutions constitute the feasible solution set. This is done to filter out hard feasible solutions that meet the actual operational constraints. Then, for each feasible solution in the feasible solution set, the comprehensive weight of each demand node is calculated and combined with the node's starting point from the scheduling origin. The weighted total delivery time is obtained by multiplying the actual delivery time to the destination and summing the results. The total transportation cost is then calculated by adding the product of the vehicle's fuel cost per unit distance and the total travel distance, plus the product of the drone's power cost per unit and the total energy consumption. This is done to quantify the timeliness and economic efficiency of the solution. Next, the sensitivity of each demand node's delivery time to fluctuations in road traffic coefficients is calculated and multiplied by the node's overall weight, then summed to obtain a robustness penalty. This is done to assess the potential impact of road condition uncertainty on the service timeliness of high-priority nodes. Finally, the weighted total delivery time, total transportation cost, and robustness penalty are multiplied by preset robustness weight coefficients and summed to obtain a comprehensive robustness index. The specific formula is as follows: In the formula, Indicates candidate solutions The comprehensive robustness index is the one that determines the best performance of the dispatching scheme in terms of timeliness, economy, and robustness to road condition fluctuations. The smaller the value of this index, the better the overall performance of the dispatching scheme in terms of timeliness, economy, and robustness to road condition fluctuations. Indicates candidate solutions The corresponding total delivery time is calculated in the same way as... Consistency is used to quantify the timeliness of the solution; Indicates candidate solutions The corresponding total transportation cost is calculated in the same way as... Consistency is used to quantify the economic performance of the plan; The weighting coefficient of the robustness penalty term is an adjustable hyperparameter used to control the degree of emphasis on delivery time fluctuations in the final scheme selection. This represents the total number of required nodes, determined by the total number of all points awaiting rescue in the emergency scenario; Indicating in candidate solutions Next, the The sensitivity of delivery time at each demand node to fluctuations in road traffic coefficients is used to assess the impact of road condition uncertainty on the service timeliness of that node.

[0058] In practical implementation, to balance computational efficiency and evaluation effectiveness, the following approach is adopted. A simplified calculation method, namely ,in, Indicates in solution In path planning, for the first The total distance traveled by vehicles or drones transporting goods to each demand node, in kilometers; This represents the average speed of the vehicle, which is a known constant, and is measured in kilometers per hour. This represents the average road traffic coefficient of the area traversed by the route, estimated from real-time traffic information; The expected fluctuation range of the road traffic coefficient is a preset constant or an estimate based on historical data. Then, the feasible solution with the smallest comprehensive robustness index is selected from the feasible solution set as the optimal scheduling scheme. This is done to achieve the best balance between timeliness, cost, and anti-interference capability, and to ensure the reliability of the final scheme in complex emergency environments.

[0059] This embodiment calculates the probability of selecting required nodes by combining comprehensive weights, spatial distances, and various environmental dynamic coefficients. An initial scheduling scheme is constructed using a step-by-step insertion method to generate an initial population. Adaptive mutation probabilities are calculated using comprehensive weights and estimated delivery times, and the population is mutated. The penalty coefficient is dynamically adjusted based on the population situation to update the model. The process iteratively solves to obtain a candidate solution set, and finally, feasible solutions are selected, with the optimal solution chosen. By prioritizing high-priority nodes, the solution efficiency and quality are improved, enabling the selection of the optimal scheduling scheme suitable for emergency scenarios and ensuring the scientific rigor and reliability of the scheduling.

[0060] Based on the first embodiment of this application, this application also provides a power material vehicle-UAV collaborative scheduling system based on node weights, please refer to... Figure 3 The system includes: The data acquisition module 10 is used to collect power grid topology data, emergency demand point information, transportation capacity resource data and real-time environmental dynamic data to obtain a basic dataset.

[0061] The weight calculation module 20 is used to calculate the comprehensive weight of each demand node based on the power grid topology data and emergency demand point information in the basic dataset.

[0062] Result module 30 is used to construct and solve a collaborative scheduling optimization model based on comprehensive weights, transportation capacity resource data and real-time environmental dynamic data to obtain the optimal scheduling scheme.

[0063] The sending module 40 is used to generate control commands based on the optimal scheduling scheme and send the control commands to the vehicle and the drone so that the vehicle and the drone can perform collaborative scheduling tasks.

[0064] The power supply vehicle-UAV collaborative scheduling system based on node weights provided in this application adopts the power supply vehicle-UAV collaborative scheduling method based on node weights in the above embodiments. It can solve the technical problem of how to integrate multi-source data to quantify node priorities and achieve joint optimization of vehicle-UAV collaborative scheduling. Compared with the prior art, the beneficial effects of the power supply vehicle-UAV collaborative scheduling system based on node weights provided in this application are the same as those of the power supply vehicle-UAV collaborative scheduling method based on node weights provided in the above embodiments. Furthermore, other technical features of the power supply vehicle-UAV collaborative scheduling system based on node weights are the same as those disclosed in the methods of the above embodiments, and will not be repeated here.

[0065] This application provides a node-weight-based power supply vehicle-UAV collaborative scheduling device. The node-weight-based power supply vehicle-UAV collaborative scheduling device includes: at least one processor; and a memory communicatively connected to the at least one processor; wherein the memory stores instructions executable by the at least one processor, and the instructions are executed by the at least one processor to enable the at least one processor to execute the node-weight-based power supply vehicle-UAV collaborative scheduling method in the above embodiment 1.

[0066] The following is for reference. Figure 4The diagram illustrates a structural schematic suitable for implementing a node-weight-based power supply vehicle-UAV collaborative scheduling device according to embodiments of this application. The node-weight-based power supply vehicle-UAV collaborative scheduling device in embodiments of this application may include, but is not limited to, mobile terminals such as mobile phones, laptops, digital broadcast receivers, PDAs (Personal Digital Assistants), PMPs (Portable Media Players), vehicle-mounted terminals (e.g., vehicle navigation terminals), and fixed terminals such as digital TVs and desktop computers. Figure 4 The power material vehicle-UAV collaborative scheduling device based on node weight shown is merely an example and should not impose any limitations on the functionality and scope of use of the embodiments of this application.

[0067] like Figure 4 As shown, the node-weight-based power supply vehicle-UAV collaborative scheduling device may include a processing unit 1001 (e.g., a central processing unit, a graphics processing unit, etc.), which can perform various appropriate actions and processes according to a program stored in a read-only memory (ROM) 1002 or a program loaded from a storage device 1003 into a random access memory (RAM) 1004. The RAM 1004 also stores various programs and data required for the operation of the node-weight-based power supply vehicle-UAV collaborative scheduling device. The processing unit 1001, ROM 1002, and RAM 1004 are interconnected via a bus 1005. An input / output (I / O) interface 1006 is also connected to the bus. Typically, the following can be connected to I / O interface 1006: input devices 1007 including, for example, touchscreens, touchpads, keyboards, mice, image sensors, microphones, accelerometers, gyroscopes, etc.; output devices 1008 including, for example, liquid crystal displays (LCDs), speakers, vibrators, etc.; storage devices 1003 including, for example, magnetic tapes, hard disks, etc.; and communication devices 1009. Communication device 1009 allows the node-weighted power supply vehicle-UAV collaborative dispatching equipment to communicate wirelessly or wiredly with other devices to exchange data. Although various node-weighted power supply vehicle-UAV collaborative dispatching devices are shown in the figure, it should be understood that implementation or possession of all shown is not required. More or fewer may be implemented alternatively.

[0068] Specifically, according to the embodiments disclosed in this application, the processes described above with reference to the flowcharts can be implemented as computer software programs. For example, embodiments disclosed in this application include a computer program product comprising a computer program carried on a computer-readable medium, the computer program containing program code for performing the methods shown in the flowcharts. In such embodiments, the computer program can be downloaded and installed from a network via a communication device, or installed from storage device 1003, or installed from ROM 1002. When the computer program is executed by processing device 1001, it performs the functions defined in the methods of the embodiments disclosed in this application.

[0069] The power supply vehicle-UAV collaborative scheduling device based on node weights provided in this application adopts the power supply vehicle-UAV collaborative scheduling method based on node weights in the above embodiments, which can solve the technical problem of how to integrate multi-source data to quantify node priorities and achieve joint optimization of vehicle-UAV collaborative scheduling. Compared with the prior art, the beneficial effects of the power supply vehicle-UAV collaborative scheduling device based on node weights provided in this application are the same as the beneficial effects of the power supply vehicle-UAV collaborative scheduling method based on node weights provided in the above embodiments, and other technical features in the power supply vehicle-UAV collaborative scheduling device based on node weights are the same as the features disclosed in the previous embodiment method, and will not be repeated here.

[0070] It should be understood that the various parts disclosed in this application can be implemented using hardware, software, firmware, or a combination thereof. In the description of the above embodiments, specific features, structures, materials, or characteristics can be combined in any suitable manner in one or more embodiments or examples.

[0071] The above description is merely a specific embodiment of this application, but the scope of protection of this application is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in this application should be included within the scope of protection of this application. Therefore, the scope of protection of this application should be determined by the scope of the claims.

[0072] This application provides a computer-readable medium having computer-readable program instructions (i.e., a computer program) stored thereon, the computer-readable program instructions being used to execute the node weight-based power material vehicle-UAV collaborative scheduling method in the above embodiments.

[0073] The computer-readable medium provided in this application may be, for example, a USB flash drive, but is not limited to electrical, magnetic, optical, electromagnetic, infrared, or semiconductor devices, or any combination thereof. More specific examples of computer-readable media may include, but are not limited to: electrical connections with one or more wires, portable computer disks, hard disks, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or flash memory), optical fibers, portable compact disk read-only memory (CD-ROM), optical storage devices, magnetic storage devices, or any suitable combination thereof. In this embodiment, the computer-readable medium may be any tangible medium containing or storing a program that can be executed by instructions, used by a device, or used in conjunction with it. The program code contained on the computer-readable medium may be transmitted using any suitable medium, including but not limited to: wires, optical cables, RF (Radio Frequency), etc., or any suitable combination thereof.

[0074] The aforementioned computer-readable medium may be included in a power material vehicle-UAV collaborative scheduling device based on node weights; or it may exist independently and not be assembled into a power material vehicle-UAV collaborative scheduling device based on node weights.

[0075] The aforementioned computer-readable medium carries one or more programs that, when executed by a node-weight-based power material vehicle-UAV collaborative scheduling device, enable the device to write computer program code for performing the operations of this application in one or more programming languages ​​or a combination thereof. These programming languages ​​include object-oriented programming languages—such as Java, Smalltalk, and C++—and conventional procedural programming languages—such as C or similar languages. The program code can be executed entirely on the user's computer, partially on the user's computer, as a standalone software package, partially on the user's computer and partially on a remote computer, or entirely on a remote computer or server. In cases involving remote computers, the remote computer can be connected to the user's computer via any type of network—including a local area network (LAN) or a wide area network (WAN)—or can be connected to an external computer (e.g., via the Internet using an Internet service provider).

[0076] The flowcharts and block diagrams in the accompanying drawings illustrate the architecture, functionality, and operation of possible implementations of methods and computer program products according to various embodiments of this application. In this regard, all blocks in the flowcharts or block diagrams may represent a module, segment, or portion of code containing one or more executable instructions for implementing the specified logical function. It should also be noted that in some alternative implementations, the functions indicated in the blocks may occur in a different order than those indicated in the drawings. For example, two consecutively indicated blocks may actually be executed substantially in parallel, and they may sometimes be executed in reverse order, depending on the functions involved. It should also be noted that all blocks in the block diagrams and / or flowcharts, and combinations of blocks in the block diagrams and / or flowcharts, may be implemented using dedicated hardware-based implementations that perform the specified functions or operations, or using a combination of dedicated hardware and computer instructions.

[0077] The modules described in the embodiments of this application can be implemented in software or hardware. The names of the modules do not necessarily limit the functionality of the unit itself.

[0078] The readable medium provided in this application is a computer-readable medium, which stores computer-readable program instructions (i.e., a computer program) for executing the above-described node-weight-based power material vehicle-UAV collaborative scheduling method. This solves the technical problem of how to integrate multi-source data to quantify node priorities and achieve joint optimization of vehicle-UAV collaborative scheduling. Compared with the prior art, the beneficial effects of the computer-readable medium provided in this application are the same as those of the node-weight-based power material vehicle-UAV collaborative scheduling method provided in the above embodiments, and will not be repeated here.

[0079] This application also provides a computer program product, including a computer program that, when executed by a processor, implements the steps of the above-described node weight-based power material vehicle-UAV collaborative scheduling method.

[0080] The computer program product provided in this application can solve the technical problem of how to integrate multi-source data to quantify node priorities and achieve joint optimization of vehicle-UAV collaborative scheduling. Compared with the prior art, the beneficial effects of the computer program product provided in this application are the same as those of the node weight-based power material vehicle-UAV collaborative scheduling method provided in the above embodiments, and will not be repeated here.

[0081] The above description is only a part of the embodiments of this application and does not limit the patent scope of this application. All equivalent structural transformations made under the technical concept of this application and using the contents of the specification and drawings of this application, or direct / indirect applications in other related technical fields, are included in the patent protection scope of this application.

Claims

1. A node weight-based power logistics vehicle-unmanned aerial vehicle collaborative scheduling method, characterized in that, include: Collect power grid topology data, emergency demand point information, transportation capacity resource data, and real-time environmental dynamic data to obtain a basic dataset; The comprehensive weight of each demand node is calculated based on the power grid topology data and emergency demand point information in the aforementioned basic dataset. The specific formula is as follows: in Indicates the first The comprehensive weight of each demand node represents its priority order in the global emergency dispatch. The larger the value, the more priority the node should receive in the dispatch scheme. Indicates the first The betweenness centrality of a demand node characterizes the importance of that node as a shortest path bridge in the power grid topology; its value is calculated from the power grid topology. This indicates the betweenness centrality among all demand nodes. The maximum value; The adjustment hyperparameter represents the balance between topological importance and real-time disaster situation and social impact, and its value ranges from [0,1]. Indicates the first Disaster severity rating of each demand node of Power of 1 Used to quantify the real-time urgency of the disaster situation in the area where the node is located, it is usually obtained from on-site assessment or remote sensing data. The nonlinear adjustment hyperparameter representing the disaster severity level is a real number greater than zero. The weighting of high-level disaster points has an amplifying effect; This represents the adjustment hyperparameter that balances the disaster severity factor and population density factor in real-time disaster information, and its value range is [value range missing]. ; Indicates the first The population density of the area associated with each demand node is used to measure the scale of social benefits that can be generated by serving that node, in units of... , This represents the population density among all demand nodes. The maximum value; This represents the index of the demand node, which is a positive integer and its value ranges from 1 to 2. This is used to uniquely identify each node that needs material assistance. This represents the total number of required nodes, determined by the total number of points awaiting rescue in the emergency scenario; Represents the hyperbolic tangent function; A collaborative scheduling optimization model is constructed and solved based on the comprehensive weight, the transportation capacity resource data, and the real-time environmental dynamic data to obtain the optimal scheduling scheme. Control commands are generated based on the optimal scheduling scheme, and the control commands are sent to the vehicle and the drone to enable the vehicle and the drone to perform collaborative scheduling tasks. The step of constructing and solving a collaborative scheduling optimization model based on the comprehensive weights, the transportation capacity resource data, and the real-time environmental dynamic data to obtain the optimal scheduling scheme includes: The selection probability of each demand node is calculated based on the comprehensive weight, spatial distance in the emergency demand point information, and road traffic coefficient, flight environment coefficient, and disaster severity level change rate in the real-time environmental dynamic data. Based on the selection probabilities, an initial scheduling scheme is constructed using a stepwise insertion method to generate an initial population; The estimated delivery time is calculated based on the path planning and task allocation of each individual in the initial population. The adaptive mutation probability is calculated based on the comprehensive weight and the estimated delivery time. The individuals in the initial population are then mutated according to the adaptive mutation probability to obtain the mutated population. The penalty coefficient is dynamically adjusted based on the proportion of feasible solutions in the population and the constraint violation of high-weight nodes, and the cooperative scheduling optimization model is updated to obtain the updated cooperative scheduling optimization model. The mutated population is iteratively solved according to the updated cooperative scheduling optimization model to obtain a candidate solution set. Based on the candidate solution set, feasible solutions that satisfy the constraints are selected, and the optimal solution is selected from the feasible solutions as the optimal scheduling scheme. The steps of calculating the estimated delivery time based on the path planning and task allocation of each individual in the initial population, calculating the adaptive mutation probability based on the comprehensive weight and the estimated delivery time, and performing mutation operations on the individuals in the initial population based on the adaptive mutation probability to obtain the mutated population include: Obtain the first comprehensive weight of the current demand node and the estimated delivery time calculated based on the path planning and task allocation of the current demand node in the initial population; and obtain the second comprehensive weight of all demand nodes and the maximum delivery time of all individuals in the initial population. The relative weight value is obtained by calculating the ratio of the maximum value of the first comprehensive weight to the maximum value of the second comprehensive weight, and the relative timeliness value is obtained by calculating the ratio of the estimated delivery time to the maximum delivery time. The relative weight value and the relative time value are multiplied to obtain the weighted time product, and the difference between the weighted time product and the variation threshold is calculated according to the preset variation threshold to obtain the variation trigger deviation; The mutation triggering deviation is transformed by a function mapping based on a preset steepness coefficient to obtain a probability adjustment coefficient; The adaptive mutation probability is calculated based on the preset minimum mutation probability, the preset maximum mutation probability, and the probability adjustment coefficient. The specific formula is as follows: in Indicates the first The adaptive mutation probability of the gene locus corresponding to each demand node; This indicates the preset minimum mutation probability; This indicates the preset maximum mutation probability. ; Represents the natural exponential function; The steepness coefficient of the Sigmoid function is represented by a hyperparameter that can be adjusted. This represents the maximum value of the overall weight across all demand nodes, used to normalize the weight factors and make the ratio... The range of values ​​is ; This indicates that in the scheduling scheme represented by the current individual, the [number]th [individual / individual]... Delivery time of materials at each demand node; This represents the maximum resource delivery time for a single individual among all individuals in the current population. It is used to normalize delivery times, making the ratio [value missing]. The range of values ​​is , This represents the threshold offset parameter of the Sigmoid function, which is an adjustable hyperparameter. Based on the adaptive mutation probability, the gene loci in the initial population related to the current demand node are mutated to obtain the mutated population.

2. The method as described in claim 1, characterized in that, The step of calculating the comprehensive weight of each demand node based on the power grid topology data and emergency demand point information in the basic dataset includes: Extract the betweenness centrality, disaster severity level, and associated population density of each demand node from the aforementioned basic dataset; The betweenness centrality is normalized to obtain a topological importance index; The severity level of the disaster is adjusted according to a preset disaster adjustment index to obtain a disaster urgency index; The social benefit index is obtained by normalizing the associated population density. The topological importance index, the disaster urgency index, and the social benefit index are weighted and summed according to a preset topological balance coefficient to obtain a comprehensive weight.

3. The method as described in claim 1, characterized in that, The step of constructing a collaborative scheduling optimization model based on the comprehensive weights, the transportation capacity resource data, and the real-time environmental dynamic data includes: The total delivery time is calculated based on the comprehensive weight and the road traffic coefficient in the real-time environmental dynamic data. The total transportation cost is calculated based on the vehicle unit distance fuel cost, vehicle total travel distance, drone unit energy consumption electricity cost, and drone total energy consumption in the aforementioned transportation resource data. Based on the aforementioned transportation resource data, the following constraints are calculated: vehicle capacity constraint penalty value, UAV range constraint penalty value, UAV payload constraint penalty value, and vehicle-UAV coordination time constraint penalty value. The total constraint violation penalty value is obtained by summing the vehicle capacity constraint penalty value, the UAV range constraint penalty value, the UAV payload constraint penalty value, and the vehicle-UAV coordination time constraint penalty value. The total transportation cost, total delivery time, and total constraint violation penalty value are linearly weighted according to preset timeliness weight coefficient, preset economy weight coefficient, and preset penalty weight coefficient to obtain the comprehensive optimization objective function value; A collaborative scheduling optimization model is constructed based on the comprehensive optimization objective function value.

4. The method as described in claim 3, characterized in that, The step of calculating the vehicle capacity constraint penalty value, the UAV range constraint penalty value, the UAV payload constraint penalty value, and the vehicle-UAV coordination time constraint penalty value based on the transportation resource data includes: Extract the maximum load capacity of the vehicle from the transportation resource data, determine the actual load weight of the vehicle, and calculate the first difference between the actual load weight and the maximum load capacity. When the first difference is greater than the preset capacity threshold, the vehicle capacity constraint penalty value is calculated based on the first difference and the preset capacity penalty coefficient. Extract the maximum range of the UAV from the transportation resource data, determine the actual flight distance of the UAV, and calculate a second difference between the actual flight distance and the maximum range; When the second difference is greater than the preset range threshold, the range constraint penalty value of the UAV is calculated based on the second difference and the preset range penalty coefficient. Extract the maximum payload of the UAV from the transportation resource data, determine the actual payload weight of the UAV, and calculate the third difference between the actual payload weight and the maximum payload. When the third difference is greater than the preset load threshold, the load constraint penalty value of the UAV is calculated based on the third difference and the preset load penalty coefficient. Extract the vehicle's stopping time at the pick-up point from the transportation resource data, determine the total mission time of the drone, and calculate the fourth difference between the total mission time and the stopping time; When the fourth difference is greater than the preset collaboration threshold, the collaboration time constraint penalty value between the vehicle and the drone is calculated based on the fourth difference and the preset collaboration penalty coefficient.

5. The method as described in claim 1, characterized in that, The step of constructing an initial scheduling scheme and generating an initial population using a step-by-step insertion method based on the selection probability includes: Obtain the set of demand nodes that are not currently being served, and randomly select a node to be served from the set of demand nodes according to the selection probability; Based on the maximum load capacity of vehicles in the transportation resource data, determine whether the node to be served is suitable for direct service by a vehicle. If it is suitable for direct service by a vehicle, insert the node to be served into the current vehicle path. If it is not suitable for direct service by a vehicle, mark the node to be served as a candidate node for drones. Based on the maximum range and maximum payload of the drone in the transportation resource data, determine whether the candidate drone node can be served by the drone from the current docking point. If it can be served by the drone from the current docking point, assign the candidate drone node to the corresponding drone task sequence. Until all demand nodes are allocated, the current scheduling scheme is constructed based on the driving paths of all vehicles and the task allocation scheme of all UAVs, and the current scheduling scheme is encoded as a population individual; Until the preset population size is reached, all individuals in the population are combined into an initial population.

6. The method as described in claim 1, characterized in that, The step of dynamically adjusting the penalty coefficient based on the proportion of feasible solutions in the population and the constraint violation of high-weight nodes to update the cooperative scheduling optimization model and obtain the updated cooperative scheduling optimization model includes: Obtain the number of individuals violating the constraints and the population size in the current population, and calculate the ratio of the number of individuals to the population size to obtain the violation ratio; Obtain a set of high-weight nodes, and obtain the weights of nodes in the set of high-weight nodes affected by vehicle overloading and the total weight of the set of high-weight nodes, wherein the set of high-weight nodes consists of nodes whose overall weight is greater than the average weight of all demand nodes. The high-weight violation rate is obtained by calculating the ratio of the node weight to the total weight. The violation ratio is linearly adjusted according to a preset linear adjustment factor to obtain the adjusted violation ratio. The high-weight violation rate is adjusted exponentially according to a preset exponential adjustment factor to obtain the adjusted high-weight violation rate. The preset initial penalty coefficient is dynamically adjusted based on the adjusted violation ratio and the adjusted high-weight violation rate to obtain the dynamic penalty coefficient. The vehicle capacity constraint penalty coefficient, UAV range constraint penalty coefficient, UAV payload constraint penalty coefficient, and vehicle-UAV coordination time constraint penalty coefficient in the collaborative scheduling optimization model are updated based on the dynamic penalty coefficient to obtain the updated collaborative scheduling optimization model.

7. The method as described in claim 1, characterized in that, The step of filtering feasible solutions that satisfy the constraints based on the candidate solution set, and selecting the optimal solution from the feasible solutions as the optimal scheduling scheme, includes: The total constraint violation penalty value for each candidate solution is calculated based on the candidate solution set. When the total constraint violation penalty value is less than the preset feasible threshold, the candidate solution is determined as a feasible solution, and all the feasible solutions are combined into a feasible solution set; For each feasible solution in the feasible solution set, the weighted total delivery time is calculated based on the comprehensive weight and the delivery time of each demand node in the feasible solution; The total transportation cost is calculated based on the fuel cost per unit distance of the vehicle, the total distance traveled by the vehicle, the unit energy cost of the drone, and the total energy consumption of the drone. A robustness penalty term is calculated based on the sensitivity of the delivery time of each demand node to fluctuations in the road traffic coefficient. The weighted total delivery time, the total transportation cost, and the robustness penalty term are weighted and summed according to the preset robustness weight coefficient to obtain a comprehensive robustness index. The feasible solution with the smallest comprehensive robustness index is selected from the feasible solution set as the optimal scheduling scheme.

8. A power material vehicle-UAV collaborative scheduling system based on node weights, characterized in that, The system is applied to the power material vehicle-UAV collaborative scheduling method based on node weights as described in any one of claims 1-7, and the system includes: The data acquisition module is used to collect power grid topology data, emergency demand point information, transportation capacity resource data, and real-time environmental dynamic data to obtain a basic dataset. The weight calculation module is used to calculate the comprehensive weight of each demand node based on the power grid topology data and emergency demand point information in the basic dataset. The results module is used to construct and solve a collaborative scheduling optimization model based on the comprehensive weight, the transportation capacity resource data, and the real-time environmental dynamic data to obtain the optimal scheduling scheme. The sending module is used to generate control commands according to the optimal scheduling scheme and send the control commands to the vehicle and the drone so that the vehicle and the drone can perform collaborative scheduling tasks.

Images