An edge-computing-based post-earthquake unmanned vehicle coordination scheduling method

Abstract

The application provides an earthquake post-unmanned vehicle and unmanned aerial vehicle cooperative scheduling method based on edge computing, belongs to the technical field of earthquake disaster rescue scheduling, and comprises the following steps: initializing an unmanned vehicle and an unmanned aerial vehicle cluster, and acquiring rescue material state data and wounded data; constructing a comprehensive target function, solving the comprehensive target function based on the rescue material state data and the wounded data, and obtaining material and wounded scheduling; acquiring required real-time data based on an edge node, feeding back the real-time data to a control center through an edge server, re-planning a path and calculating unloading, and dynamically allocating the material and wounded scheduling; and improving post-earthquake material transfer timeliness and wounded survival rate by using intelligent scheduling of unmanned vehicle and unmanned aerial vehicle cooperation.

Description

Technical Field

[0001] This invention belongs to the field of earthquake disaster relief and dispatch technology, specifically involving a post-earthquake unmanned vehicle-machine collaborative dispatch method based on edge computing. Background Technology

[0002] In earthquake disaster relief scenarios, traditional material transportation and casualty evacuation face multiple challenges, including road damage, communication disruptions, and frequent secondary disasters, limiting rescue efficiency. Drones and unmanned vehicles (UAVs) achieve real-time communication through dynamic heterogeneous networks. Their onboard high-definition cameras and the UAVs' lidar form a three-dimensional perception system, capable of constructing centimeter-level precision 3D disaster area maps, providing a foundation for route planning. Existing invention patent "Unmanned Vehicle System and Control Method for Earthquake Disaster Relief" (CN202410316846.X) falls under the field of earthquake disaster relief technology. It is a single-vehicle UAV rescue system, lacking UAV collaboration and edge computing support, resulting in limited dynamic response capabilities. Invention patent "Disaster Relief Scheduling Method, Device, and System Based on Edge Computing" (CN201811003618.8) involves a disaster relief scheduling method, device, and system based on edge computing, but it does not consider multi-vehicle collaboration and is not specifically designed for post-earthquake scenarios.

[0003] Therefore, existing unmanned vehicle-machine collaborative scheduling technologies have certain limitations and cannot cope with the rapidly changing information after a disaster, nor can they simultaneously integrate the two-way tasks of material delivery and casualty transfer. They cannot solve the problem of multi-objective dynamic optimization of relief material and casualty transfer through unmanned vehicle-machine collaborative scheduling under conditions of road damage and communication restrictions after an earthquake. Summary of the Invention

[0004] To address the limitations of existing technologies in the collaborative scheduling of unmanned vehicles and drones in actual post-earthquake scenarios, this invention provides a post-earthquake unmanned vehicle and drone collaborative scheduling method based on edge computing.

[0005] To achieve the above objectives, the present invention provides the following technical solution:

[0006] A post-earthquake unmanned vehicle-machine collaborative scheduling method based on edge computing includes the following steps:

[0007] Initialize an unmanned vehicle and drone cluster to acquire status data of relief supplies and casualty data; construct a comprehensive objective function with the goals of minimizing total rescue time, maximizing safety, minimizing total energy consumption, ensuring fairness in casualty transfer, and minimizing communication latency.

[0008] The comprehensive objective function is solved based on the status data of the relief supplies and the data of the wounded to obtain the scheduling plan for the supplies and the wounded.

[0009] In the event of aftershocks, roadside units with edge servers are used as edge nodes. Real-time road condition changes, vehicle status, fluctuations in material demand, and new casualty data are obtained based on these edge nodes. The real-time road condition changes, vehicle status, fluctuations in material demand, and new casualty data are fed back to the control center through the edge servers for re-planning and recalculation of unloading, and redistribution of material and casualty dispatch.

[0010] Preferably, the comprehensive objective function is as follows:

[0011] ;

[0012] in, This is the maximum time required to complete the rescue operation. For relief supplies At the receiving point Emergency weight; This refers to positive and negative deviations in material allocation. For receiving point relief supplies Demand; Type of wounded From the receiving point Average arrival time to the rescue station; It is the time threshold for assessing the urgency of the injured. yes Priority weights for different types of casualties; These represent the total energy consumption of the driverless car and the drone, respectively. This represents the total energy consumption and total communication latency of edge computing.

[0013] Preferably, the method further includes setting constraints on the comprehensive objective function, including:

[0014] Specifically, when the total rescue time is minimized: when material or casualty handover occurs, the time difference between the arrival of unmanned vehicles and drones at the transfer point must be within a set value; the time for unmanned vehicles or drones to leave the rescue station must not be less than the sum of their arrival time and the time for loading rescue materials, and the time for unmanned vehicles or drones to leave the receiving point must not be less than the sum of their arrival time, the time for unloading rescue materials, and the time for picking up casualties, and loading casualties can only begin after the unloading of materials is completed; set the final time for all vehicles to complete their tasks.

[0015] To maximize safety, the following constraints should be set: the payload of unmanned vehicles and drones should not exceed the capacity weight limit; the maximum battery capacity of drones should limit their maximum range; and the total data transmission volume of a single communication link should not exceed the bandwidth.

[0016] To minimize total energy consumption, the energy consumption of unmanned vehicles, drones, and edge computing should not exceed the set energy consumption; the maximum battery power of drones limits their maximum range.

[0017] In ensuring fairness in the transfer of the wounded, it is stipulated that the wounded in special emergencies must be directly transported by drones, and the transfer of the wounded in very urgent situations is prohibited. The wounded in general are handed over by drones to unmanned vehicles at the transfer point or transported by unmanned vehicles from the receiving point to the rescue station by unmanned vehicles. The unmanned vehicles and drones are subject to time and space synchronization constraints.

[0018] To minimize communication latency, each computing task must be assigned to an edge server for processing upon generation; the total latency, consisting of transmission latency and computing latency, must not exceed the maximum allowable latency; and the total computing load of a single edge server must not exceed its computing power limit in real time.

[0019] Preferably, the comprehensive objective function is solved based on the status data of the relief supplies and the data of the wounded to obtain the scheduling plan for the supplies and the wounded. Specifically, the Benders decomposition method is used to solve the problem.

[0020] Preferably, the Benders decomposition method decomposes the problem into a main problem and sub-problems. The main problem is a material allocation and casualty evacuation plan, and the sub-problems are path planning and edge computing offloading. CPLEX is used to solve the linear programming main problem, while Gurobi is used to solve the mixed integer programming sub-problems, thus obtaining the material and casualty scheduling plan.

[0021] Preferably, the unmanned vehicles communicate with each other via V2V communication technology to achieve information exchange between vehicles.

[0022] This invention also provides a post-earthquake unmanned vehicle-machine collaborative scheduling system based on edge computing, specifically including:

[0023] The initialization module is used to initialize the unmanned vehicle and drone clusters, and to acquire data on the status of rescue supplies and casualties. A comprehensive objective function is constructed with the goals of minimizing total rescue time, maximizing safety, minimizing total energy consumption, ensuring fairness in casualty transfer, and minimizing communication latency.

[0024] The objective optimization module solves the comprehensive objective function based on the status data of the relief supplies and the data of the wounded, and obtains the scheduling plan for the supplies and the wounded.

[0025] The dynamic scheduling module is used to, in the event of aftershocks, use roadside units with edge servers as edge nodes to acquire real-time data on road conditions, vehicle status, fluctuations in material demand, and new casualties. The real-time data on road conditions, vehicle status, fluctuations in material demand, and new casualties is fed back to the control center through the edge servers for re-planning and recalculation of routes and unloading, and for redistribution of material and casualty scheduling.

[0026] The present invention also provides a computer device, including a memory, a processor, and a computer program stored in the memory, wherein the processor executes the computer program to implement the steps described in the post-earthquake unmanned vehicle-machine collaborative scheduling method based on edge computing.

[0027] The present invention also provides a computer-readable storage medium storing a computer program, which, when loaded by a processor, is capable of executing the steps described in the post-earthquake unmanned vehicle-machine collaborative scheduling method based on edge computing.

[0028] The post-earthquake unmanned vehicle-machine collaborative scheduling method based on edge computing provided by this invention has the following beneficial effects:

[0029] This invention utilizes the planning and scheduling of unmanned vehicles and drones to rapidly and accurately transport relief supplies from rescue stations to receiving points (disaster sites), and safely transfer the injured back to the rescue station during the return journey. Throughout the process, it fully considers various factors such as post-earthquake road damage, the classification of relief supplies and the injured, and the type of transport vehicles to achieve fairness in the allocation of relief supplies, minimize total rescue time, and maximize the efficiency of the transfer of the injured. Edge computing is introduced to enable collaborative transportation of relief supplies and the injured, overcoming efficiency bottlenecks in situations lacking edge computing capabilities. Edge computing also enables a closed loop of "perception-decision-control," overcoming the problems caused by altered information regarding the on-site situation due to road damage during aftershocks. This provides a more comprehensive and efficient solution for post-earthquake relief. Collaborative transportation addresses the coordination of multiple vehicles, edge computing handles dynamic response, and specific scenario constraints reflect the needs of earthquake relief. Attached Figure Description

[0030] To more clearly illustrate the embodiments and design schemes of the present invention, the accompanying drawings required for this embodiment will be briefly described below. The drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0031] Figure 1 This is a flowchart of a post-earthquake unmanned vehicle-machine collaborative scheduling method based on edge computing, according to the present invention. Detailed Implementation

[0032] To enable those skilled in the art to better understand and implement the technical solutions of the present invention, the present invention will be described in detail below with reference to the accompanying drawings and specific embodiments. The following embodiments are only used to more clearly illustrate the technical solutions of the present invention and should not be construed as limiting the scope of protection of the present invention.

[0033] Example

[0034] Road and airway network definition: In the complex environment following an earthquake, to achieve efficient transportation of relief supplies and evacuation of the injured, the routes for unmanned vehicles (UAVs) and drones are clearly defined. UAVs can only travel on the normal road network, but earthquakes damage some roads, rendering them unusable for UAVs; drones, on the other hand, fly on a specially defined airway network. A specific set of coupling nodes exists between the airway network and the normal road network. This set of nodes includes receiving points (i.e., disaster sites), relief stations, and transfer points. These coupling nodes play a crucial connecting role in the collaborative operation of UAVs and drones, ensuring the smooth progress of relief supply transportation and evacuation of the injured.

[0035] Direct Flight Rules: Direct drone flights refer to drones flying independently throughout the mission, without being carried by unmanned vehicles (UAVs) or having their supplies handed over to UAVs at transit points. UAVs fly directly from the origin to the destination. Direct UAV delivery refers to UAVs transporting supplies directly to the receiving point without having their supplies handed over to drones at transit points. Whether using direct drone flights or direct UAV delivery, after passing through a transit point once, direct delivery or direct flight is used, and the shortest route is selected to achieve high efficiency in transporting relief supplies and transferring the wounded, saving time and resources to the greatest extent possible.

[0036] The relief supplies are categorized into two main types: emergency medical supplies and survival supplies, to meet different levels and types of rescue needs. The injured are categorized into extremely urgent and general urgent cases to allow for appropriate transfer strategies based on the severity of their injuries. Drones are divided into heavy and light types, each with different performance characteristics in terms of payload and endurance, suitable for various transport missions. Furthermore, the total number of supply categories, receiving points, rescue stations, unmanned vehicles, and drones are all set to fixed values. This framework guides subsequent mission planning, resource allocation, and model building, ensuring clear boundaries and operability for research and technological applications.

[0037] This invention provides a post-earthquake unmanned vehicle-machine collaborative scheduling method based on edge computing, such as... Figure 1 As shown, the specific steps include:

[0038] Step 1: Construct a normal road network and a drone flight path network for unmanned vehicles. The normal road network for unmanned vehicles is as follows: , ,in, For the rescue station, As the receiving point, As a transit point, It is a road intersection; Mark the accessible arc sections (normal roads) for unmanned vehicles and the damaged sections (where unmanned vehicles are prohibited) for them. Construct an unmanned aerial vehicle (UAV) route network. , , For unmanned aerial vehicle waypoints (including temporary take-off and landing points); This represents the drone's flight arc (straight path). (Setting) For the damaged road sections ( ).

[0039] Step 2: Based on the road network and UAV route network, deploy roadside units (RSUs) with edge servers in the road network to build edge computing nodes.

[0040] Both unmanned vehicles (V2V) and drones are equipped with V2X (Vehicle-to-Everything) systems, a key technology for enabling information exchange between vehicles and the outside world. V2V (Vehicle-to-Vehicle) communication allows V2V to exchange information, such as sharing real-time vehicle status and location information. V2V also communicates wirelessly with roadside units equipped with edge computing servers. This allows V2V to transmit its collected data to the edge computing server for processing and receive computational results and instructions from the server, achieving real-time data transmission. Furthermore, drones communicate with V2V to share crucial information such as when, how much, and where of relief supplies, and when, and where, injured personnel were transferred. This real-time exchange of information provides vital data support for the coordinated scheduling of V2V and drones, relief supply delivery, and patient transfer, enabling the entire rescue operation to proceed more efficiently and systematically.

[0041] In road networks, roadside units (equipped with edge servers) are installed. These edge computing servers provide real-time computing services for devices such as autonomous vehicles (RVs). However, the network computing resources of edge computing servers are limited, while RVs requiring computing services are latency-sensitive (with millisecond-level reaction times), and assuming they cannot perform calculations themselves, their computing tasks must be distributed to edge computing servers. To reduce latency and balance the computing load of edge nodes, a matching optimization between RV computing tasks and computing resources is constructed with the goal of minimizing the total latency of the information network. Specifically, through the analysis of RV computing tasks and real-time monitoring of edge computing server resource status, reasonable algorithms and strategies are adopted to allocate RV computing tasks to the most suitable edge computing nodes. In this way, edge computing servers can reduce RV data latency by balancing the computing load of edge computing nodes. The reduction in data latency will in turn affect the RV's path-making decisions, enabling the RV to react more accurately and promptly during operation, thus improving the optimization effect of RV dispatching and delivery routes. Meanwhile, the drones adjust their paths in a timely manner through communication with the central cloud and coordinate with unmanned vehicles for optimization. Throughout the rescue process, the drones and unmanned vehicles cooperate with each other, dynamically adjusting their respective task execution methods and path planning based on real-time communication and calculation results, so as to achieve efficient completion of the rescue mission.

[0042] Step 3: Initialize the unmanned vehicle (UGV) and unmanned aerial vehicle (UAV) cluster, load the categories and inventory of relief supplies, and enter the categories and distribution locations of the injured.

[0043] Real-time data acquisition, through the V2X communication system, obtains dynamic information including: changes in road conditions, vehicle status, fluctuations in material demand, and new casualties.

[0044] set up and They are respectively groups of unmanned vehicles and groups of humans and machines; settings and For the set of material types ( =Emergency medical care, =Survival supplies) and the set of wounded types ( =Very urgent =General Emergency); Settings It is a collection of edge servers (deployed in roadside units). This refers to the set of computational tasks for autonomous vehicles (path planning, obstacle avoidance, etc.). It is a set of communication links (V2V, V2I).

[0045] Step 4: Construct a comprehensive objective function with the goals of minimizing total rescue time, maximizing safety, minimizing total energy consumption, ensuring fairness in casualty transport, and minimizing communication latency. Specifically:

[0046] ;

[0047] in, This is the maximum time required to complete the rescue operation. For relief supplies At the receiving point Emergency weight; Positive (negative) deviations in material allocation; For receiving point relief supplies Demand; Type of wounded From the receiving point Average arrival time to the rescue station; It is the time threshold for assessing the urgency of the injured. yes Priority weight of type of wounded ( ); The values ​​represent the total energy consumption (J) of the unmanned vehicle and the unmanned drone, respectively. Represents the total energy consumption of edge computing (J) and the total communication latency (ms); , .

[0048] In the objective function, the second term represents overall safety, the third term represents fairness in the transfer of wounded, the fourth term represents total energy consumption, and the fifth term represents total communication delay.

[0049] Step 5: Construct a constraint system based on the comprehensive objective function obtained in Step 4. This includes constraints on road network traffic, vehicle flow conservation, material flow conservation, casualty flow conservation, coordinated transportation, capacity and resource constraints, auxiliary constraints, communication and computing constraints, energy consumption constraints, and time continuity and consistency constraints.

[0050] The correspondence between the overall objective function and the constraints:

[0051] The constraints corresponding to the goal of minimizing the total rescue time include: spatiotemporal synchronization constraints for material handover, spatiotemporal synchronization constraints for casualty handover, chain constraints for node operation time, and global time constraints, to ensure seamless handover operations and avoid waiting delays; edge computing task time integration constraints accelerate decision-making and shorten response time.

[0052] The constraints corresponding to the goal of maximizing safety include: unmanned vehicle load constraints, drone load constraints, drone endurance constraints, and communication link capacity constraints, to prevent overload overturning / crash and to avoid control failures caused by communication congestion.

[0053] The constraints corresponding to the goal of fairness in the transfer of wounded include: wounded priority constraints, spatiotemporal synchronization constraints in the handover of wounded, and constraints on the calculation of wounded transfer time. These constraints ensure that wounded in extremely urgent situations are transported directly, and the fairness of transfer at each receiving point is quantified by controlling the transfer time of wounded through time windows.

[0054] The constraints corresponding to the goal of minimizing total energy consumption include: energy consumption constraints, drone endurance constraints, optimization paths, and task offloading strategies.

[0055] The constraints corresponding to the goal of minimizing communication latency include: communication and computation constraints to reduce processing latency, and communication link capacity constraints to ensure real-time transmission.

[0056] Road network traffic constraints include: road network constraints for unmanned vehicles, flight path constraints for drones, and restrictions on passage through damaged road sections.

[0057] The road network constraints for autonomous vehicles are as follows:

[0058] ;

[0059] in, Indicates driverless car Whether from arrive ( This formula indicates that each autonomous vehicle can only choose one exit arc at any given time and at any given node, ensuring path uniqueness.

[0060] ;

[0061] The above formula indicates that driverless vehicles can only travel on normal roads.

[0062] The specific flight path constraints for drones are as follows:

[0063] ;

[0064] in, Indicates drone Whether from arrive ( This formula indicates that a drone can only have one exit arc at any given time: each drone can only choose one exit arc at any given time and at any given node, ensuring the uniqueness of the flight path.

[0065] ;

[0066] The above formula indicates that the drone can only fly along a preset route.

[0067] The damaged road sections are closed to traffic as follows:

[0068] ;

[0069] The above formula indicates that driverless vehicles are prohibited from passing through damaged road sections.

[0070] Vehicle flow conservation constraints include: flow conservation constraints for unmanned vehicle nodes and flow constraints for unmanned vehicle vehicles.

[0071] The flow conservation constraint for autonomous vehicle nodes is as follows:

[0072]

[0073] The first equation indicates that each autonomous vehicle must depart from a certain rescue station; the second equation indicates that each vehicle must eventually return to a certain rescue station; the third equation indicates that at ordinary nodes that are not rescue stations / receiving points, the inflow of autonomous vehicles equals the outflow.

[0074] Unmanned aerial vehicle (UAV) constrained flow constraints specifically include:

[0075] The constraint requiring drones to depart from and return to the rescue station means that all drones must depart from the rescue station:

[0076] ;

[0077] All drones must eventually be returned to the rescue station. .

[0078] Traffic conservation constraints for drone nodes:

[0079] ;

[0080] in, Indicates drone Whether it is installed in driverless cars The left side of the equation represents inflow: the first term is arrival by autonomous flight, and the second term is arrival by passenger transport; the right side represents outflow: the first term is departure by autonomous flight, and the second term is departure by passenger transport.

[0081] The drone is subject to state constraints, meaning it can only be mounted on a single vehicle at most:

[0082] ;

[0083] Drones can only be carried when the driverless car is in operation. .

[0084] Material flow conservation constraints include: material flow constraints for unmanned vehicles and material flow constraints for drones.

[0085] The material flow constraints for unmanned vehicles are as follows:

[0086] ;

[0087] in, It's an autonomous vehicle. In the arc segment Transporting goods The amount; It is delivered by driverless car Point of supply quantity; It is a rescue station relief supplies Inventory levels; Indicates in Driverless cars Should we send drones? Transfer of materials; Indicates the quantity of materials transferred; It's an autonomous vehicle. In the path Transporting supplies The amount; Indicates whether the driverless car is on the path There is no handover throughout the entire process. This constraint means that the goods are unloaded at the receiving point, loaded at the rescue station, handed over to the drone at the transfer point, or delivered directly to the receiving point by the unmanned vehicle without transfer.

[0088] The constraints on drone material flow are as follows:

[0089] ;

[0090] in, Indicates drone In the arc segment Transporting goods The amount; Indicates delivery by drone supplies quantity; Indicates that the drone departed from the rescue station Loaded supplies quantity; Indicates drone In the path Transporting supplies The amount; Indicates whether the drone is on the path There is no handover throughout the entire process. This constraint means that the drone unloads at the receiving point, receives supplies at the transit point, loads at the rescue station, and flies directly from the rescue station to the receiving point without handover.

[0091] The casualty flow conservation constraints include: casualty origin flow constraints, casualty destination flow constraints, casualty flow constraints for unmanned vehicles, casualty flow constraints for drones, and global casualty conservation constraints.

[0092] The flow constraints at the origin of the wounded are as follows:

[0093] ;

[0094] in, and They represent drones and driverless cars In the arc segment transportation Number of casualties; It is a receiving point of Number of wounded of each type. This formula indicates that all wounded must be transported out of the receiving point.

[0095] The casualty destination flow constraint means that all casualties must be transported to the rescue station, specifically:

[0096] ;

[0097] The constraints on casualty flow in unmanned vehicles are as follows:

[0098] ;

[0099] in, Indicates whether it is in the node drones Towards driverless cars Handing over the wounded. This formula represents the conservation of wounded personnel at non-transfer points and the receipt of wounded personnel by drones at transfer points.

[0100] The constraints on drone casualty flow are as follows:

[0101] ;

[0102] in, Indicates at the receiving point The number of wounded soldiers on board; Indicates a transit point Handed over to driverless vehicle The number of wounded; Indicates drone In the path The number of wounded transported. This constraint indicates whether the wounded are loaded at the receiving point, transferred at the transit point, or flown directly from the receiving point to the rescue station.

[0103] The global conservation constraint for casualties states that the total number of casualties loaded is equal to the sum of the number of unmanned vehicles transferred and the number of vehicles directly transported to rescue stations, specifically:

[0104] ;

[0105] Coordinated transfer constraints include: spatiotemporal synchronization constraints for material handover, spatiotemporal synchronization constraints for casualty handover, and route verification constraints without handover.

[0106] The spatiotemporal synchronization constraints for material handover are as follows:

[0107] ;

[0108] in, Indicates driverless car drones Reaching the node Time; This is the maximum time window (min) for collaborative transport. This constraint means that when material handover or casualty handover occurs, the time difference between the arrival of the unmanned vehicle and the drone at the transfer point must be within the tolerance window. Inside.

[0109] The spatiotemporal synchronization constraint for casualty handover means that the unmanned vehicle and the drone arrive at the handover point simultaneously, specifically:

[0110] ;

[0111] The path has no intersection verification constraints, specifically:

[0112] ;

[0113] ;

[0114] in, This is a sufficiently large constant. This constraint ensures that the unmanned vehicle and the drone have no intersection path and no collaborative operation throughout the entire process.

[0115] Capability and resource constraints include: unmanned vehicle payload constraints, drone payload constraints, drone endurance constraints, and relief supply constraints.

[0116] The load constraints for autonomous vehicles are as follows:

[0117] ;

[0118] in, It is the average weight of the wounded (kg / person); It's an autonomous vehicle. Load capacity (kg); This constraint indicates that the weight of the driverless vehicle does not exceed its capacity.

[0119] The payload constraints for drones are as follows:

[0120] ;

[0121] in, It is a drone Load capacity (kg); This constraint states that the drone's payload cannot exceed its capacity.

[0122] The specific limitations of drone battery life are as follows:

[0123] ;

[0124] in, It is the arc segment of the drone Flight distance; It is a drone The maximum range (km). This constraint indicates that the drone's maximum battery capacity limits its maximum range.

[0125] Relief supply constraints mean that the amount of supplies allocated cannot exceed the supply limit, specifically:

[0126] .

[0127] Auxiliary constraints include: nonnegativity and integer constraints, and casualty priority constraints.

[0128] Nonnegativity and integer constraints, specifically:

[0129] ;

[0130] The casualty priority constraint means that critically injured personnel must be transported directly by drone, specifically:

[0131] ;

[0132] The transfer of critically injured patients is prohibited: .

[0133] Generally, at transfer points, the wounded are handed over by drones to unmanned vehicles, or transported from the receiving point to the rescue station by unmanned vehicles. ;

[0134] in, Representation type Threshold for the proportion of wounded soldiers directly transported. This constraint represents...

[0135] Communication and computing constraints include: task offloading constraints, edge computing power constraints, latency constraints, and communication link capacity constraints.

[0136] Task unloading constraints are as follows:

[0137] ;

[0138] ;

[0139] in, Indicates task Should it be offloaded to the edge server? Processing. This constraint indicates that when an autonomous vehicle... Reaching the edge server The node where the computation task is located will generate a computation task. Each computational task During generation, it must be assigned to an edge server for processing (because the autonomous vehicle has no local computing power); this satisfies the assumption of "full offloading of computing tasks" for the autonomous vehicle.

[0140] Edge computing power constraints are as follows:

[0141] ;

[0142] in, It is an edge server CPU computing power (GHz). This constraint represents the CPU computing power of a single edge server. The total computing load must not exceed its computing power limit in real time; edge servers may be damaged after an earthquake, so overload must be strictly avoided.

[0143] The delay constraint is as follows:

[0144]

[0145] in, It is a task Data volume (MB); It is a task The calculated intensity; It is a link The bandwidth (Mbps); Indicates task Should a link be used? transmission; It is a task Maximum allowable delay (ms). This constraint states that the total delay, consisting of transmission delay and computation delay, does not exceed the maximum allowable delay.

[0146] The communication link capacity constraints are as follows:

[0147]

[0148] This constraint means that the total data transmission volume of a single communication link (V2V / V2I) cannot exceed the bandwidth; aftershocks may cause link instability, so bandwidth redundancy needs to be reserved.

[0149] Energy consumption constraints include: energy consumption of autonomous vehicles, energy consumption of drones, and energy consumption of edge computing.

[0150] The energy consumption of autonomous vehicles is as follows:

[0151] ;

[0152] in, The driverless car is on the arc. The driving distance; It's an autonomous vehicle. Energy consumption per unit distance of travel (J / km); It is a link Energy consumption per unit of data transmission (J / MB). In the above formula, the first term represents the energy consumption of the autonomous vehicle on the go, and the second term represents the energy consumption of the autonomous vehicle's edge computing and communication.

[0153] The specific energy consumption of drones is as follows:

[0154] ;

[0155] in, It is a drone Energy consumption per unit distance during flight (J / km). In the above formula, the first term represents the energy consumption of the UAV during flight, and the second term represents the energy consumption of the UAV's edge computing and communication.

[0156] The energy consumption of edge computing is as follows:

[0157] ;

[0158] in, Represents edge server Unit computing power consumption (J / GHz). This refers to the power consumption of the roadside edge computing system; optimizing this power consumption is significant in encouraging the offloading of computing tasks to energy-efficient edge servers.

[0159] Time continuity and consistency constraints include: node operation time chain constraints, arrival time and arc segment time correlation constraints, casualty transfer time calculation constraints, computation task time integration constraints, and global time constraints.

[0160] Node operation time chain constraint:

[0161] (1) Rescue station departure time constraints, specifically:

[0162] ;

[0163] ;

[0164] ;

[0165] ;

[0166] in, Indicates driverless car drones Leave node Time; It is used to load relief supplies. The unit time (min / unit). This constraint means that the time it takes for the unmanned vehicle (drone) to leave the rescue station is not less than the sum of its arrival time and the time it takes to load rescue supplies.

[0167] (2) Receiving point departure time constraint, specifically:

[0168] ;

[0169] ;

[0170] ;

[0171] ;

[0172] in, It is unloading relief supplies Unit time (min / unit); It is for transporting the wounded. The unit time (min / person). This constraint means that the time for the unmanned vehicle (drone) to leave the receiving point is not less than the sum of its arrival time, the time for unloading relief supplies, and the time for picking up the wounded. Ensure that the previous operation is completed before starting the next operation (e.g., unloading of supplies must be completed before loading of the wounded can begin).

[0173] (3) Departure time constraints at transit points, specifically:

[0174] ;

[0175] ;

[0176] ;

[0177] in, It is the operation time for a single handover of relief supplies (min / time); This indicates the start time of the unmanned vehicle / drone handover. This constraint indicates the departure time after the handover of supplies.

[0178] ;

[0179] ;

[0180] ;

[0181] in, This is the operation time (min / transaction) for a single casualty handover. This constraint indicates the departure time after the casualty handover; the time for the unmanned vehicle (drone) to leave the receiving point must not be less than its handover end time. Collaborative operations can only begin after both parties have arrived.

[0182] The constraint relating arrival time to arc time is as follows:

[0183] ;

[0184] .

[0185] The constraints for calculating the time required for transporting the wounded are as follows:

[0186] ;

[0187] This constraint indicates the injured person From the receiving point The average time to reach the rescue station, with the numerator being the total waiting time of the wounded and the denominator being the total number of wounded.

[0188] The computational task time integration constraints are as follows:

[0189] ;

[0190] ;

[0191] in, It is a task On the edge server Processing time (ms) This constraint indicates that the autonomous vehicle... Uninstallation task Upon reaching the edge server node Then begin the calculation.

[0192] Task You can only leave the edge server node after completing the task:

[0193] ;

[0194] This constraint means that the autonomous vehicle is triggered after it arrives at the edge server node (task unloading, task calculation, path planning, etc.); the next path segment can only be determined after the calculation is completed.

[0195] The global time constraints are as follows:

[0196] ;

[0197] This constraint represents the final time when all vehicles complete their missions. As a decision variable, it enhances scheduling flexibility.

[0198] Step Six: The problem is decomposed into a main problem and subproblems using the Benders decomposition method. The main problem is the allocation of supplies and the transfer of the wounded, while the subproblems are path planning and computational unloading. Therefore, CPLEX is used to solve the linear programming main problem, while Gurobi is used to solve the mixed-integer programming subproblems, thus obtaining the optimal collaborative solution for unmanned vehicles and drones based on edge computing for the transfer of relief supplies and the wounded during post-earthquake rescue. The specific steps of solving this problem using the Benders decomposition method are as follows:

[0199] First, the problem is decomposed into a main problem and sub-problems. The main problem is resource allocation and casualty evacuation plans, while the sub-problems are path planning and computational offloading. The introduction of edge computing tasks means that the main problem requires simultaneous decisions on resource allocation and task offloading, while the sub-problems need to verify the feasibility of coordinating communication, computing, and logistics under given decisions.

[0200] The main problem (MP) is a resource allocation and task offloading problem. Decision variables include the amount of supplies allocated, the number of wounded personnel allocated, the decision on task offloading, and channel allocation. The objective is to minimize the weighted sum of the total rescue time and the total task processing delay. The constraints include those directly related to resource allocation and task offloading: material supply constraints, vehicle capacity constraints, computing capacity, casualty priority constraints, computing task requirements constraints, communication resource constraints, and Benders cut constraints. The output is an integrated solution, including decisions on material allocation, casualty evacuation, and computing task offloading. .

[0201] Subproblem (SP) - Feasibility verification problem of integrated cooperative scheduling (solution to the fixed main problem) Decision variables include path variables, time variables, flow variables, and handover state variables; the objective is to verify the feasibility of a fixed master problem solution, and if feasible, to calculate its actual total time cost. (Integrates rescue time and mission delay); constraints are all relevant constraints of the original model; outputs the feasibility state, and if feasible, outputs the actual total rescue time. And used to generate dual information for cutting.

[0202] The algorithm iterative steps are as follows:

[0203] (1) Initialization: Set the convergence tolerance ε>0, set the upper bound UB=+∞, set the lower bound LB=-∞, and set the iteration counter k=0.

[0204] (2) Solve the main problem MP:

[0205] Solving for MP yields an integrated solution. and its target value .

[0206] Update the lower bound: LB= LB is a theoretical lower limit for the total rescue time of a system under ideal conditions that ignore complex spatiotemporal coordination, communication fluctuations, and computational queuing.

[0207] (3) Solve the subproblem SP k The solution to the main problem Substitute it into the subproblem as a fixed parameter.

[0208] If the subproblems are feasible and the actual total time cost is calculated. .

[0209] Update the upper bound: UB = min(UB, T), where UB is the optimal total system time cost (rescue time + mission delay) that can be achieved in reality up to this iteration. Generate an optimal cut. And add it to the main question.

[0210] If the subproblem is infeasible, generate a feasible cut. (This is a linear constraint on the variables of the main problem), and it is added to the main problem to exclude solutions that lead to infeasibility.

[0211] (4) Convergence check:

[0212] If UB-LB≤ε, the algorithm terminates; otherwise, let k=k+1 and return to (2) to continue the iteration.

[0213] (5) Result Output. The optimal solution is the integration scheme corresponding to UB. and scheduling scheme .

[0214] Using Benders decomposition to solve the problem reduces the complexity of the solution. For the first time, a weighted objective function integrates rescue time, safety, fairness, energy consumption, and communication latency into a unified framework, achieving multi-objective collaborative optimization. Edge computing is introduced to enable collaborative transportation of rescue supplies and the wounded, breaking through the efficiency bottleneck in the absence of edge computing. Edge computing enables a closed loop of "perception-decision-control," overcoming problems such as changes in road network topology caused by aftershocks and information silos caused by communication fluctuations for unmanned vehicles.

[0215] This invention also provides a post-earthquake unmanned vehicle-machine collaborative scheduling system based on edge computing, comprising:

[0216] The initialization module is used to initialize the unmanned vehicle and drone clusters, and to acquire data on the status of rescue supplies and casualties. A comprehensive objective function is constructed with the goals of minimizing total rescue time, maximizing safety, minimizing total energy consumption, ensuring fairness in casualty transfer, and minimizing communication latency.

[0217] The objective optimization module solves the comprehensive objective function based on the status data of relief supplies and the data of the wounded, and obtains the scheduling plan for supplies and the wounded.

[0218] The dynamic scheduling module is used to, in the event of aftershocks, use roadside units with edge servers as edge nodes to obtain real-time data on road conditions, vehicle status, fluctuations in material demand, and new casualties. The real-time data on road conditions, vehicle status, fluctuations in material demand, and new casualties is fed back to the control center through the edge servers for re-planning and recalculation of routes and unloading, and redistribution of material and casualty dispatch.

[0219] The modules in the aforementioned post-earthquake unmanned vehicle-machine collaborative scheduling system based on edge computing can be implemented entirely or partially through software, hardware, or a combination thereof. These modules can be embedded in or independent of the processor in a computer device, or stored in the memory of a computer device as software, so that the processor can call and execute the corresponding operations of each module.

[0220] The present invention also provides a computer device, including a memory, a processor, and a computer program stored in the memory. The processor executes the computer program to implement the steps in an embodiment of a post-earthquake unmanned vehicle-machine collaborative scheduling method based on edge computing. Specific implementation methods can be found in the method embodiments, and will not be repeated here.

[0221] Furthermore, the present invention also provides a non-transitory computer-readable storage medium containing instructions on which a computer program is stored. For example, a memory containing instructions that can be executed by a processor of a computer device to complete the above-described method. For example, the non-transitory computer-readable storage medium can be a ROM, random access memory (RAM), CD-ROM, magnetic tape, floppy disk, and optical data storage device, etc. When the computer program is executed by the processor, it can implement the steps in an embodiment of a post-earthquake unmanned vehicle-machine cooperative scheduling method based on edge computing. Specific implementation methods can be found in the method embodiments, which will not be repeated here.

[0222] Those skilled in the art will understand that embodiments of the present invention can provide methods, systems, or computer program products. Therefore, the present invention can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, the present invention can take the form of a computer program product embodied on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0223] This invention is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of the invention. It will be understood that each block of the flowchart illustrations and / or block diagrams, as well as combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, generate instructions for implementing the flowchart... Figure 1 One or more processes and / or boxes Figure 1 A device that provides the functions specified in one or more boxes.

[0224] These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing device to function in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means, which are implemented in a process Figure 1 One or more processes and / or boxes Figure 1 The function specified in one or more boxes.

[0225] These computer program instructions may also be loaded onto a computer or other programmable data processing equipment to cause a series of operational steps to be performed on the computer or other programmable equipment to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable equipment for implementing the process. Figure 1 One or more processes and / or boxes Figure 1 The steps of the function specified in one or more boxes.

[0226] It should be noted that the specific embodiments described above enable those skilled in the art to more fully understand the present invention, but do not limit the present invention in any way. Therefore, although the present invention has been described in detail in this specification and embodiments, those skilled in the art should understand that modifications or equivalent substitutions can still be made to the present invention; and all technical solutions and improvements that do not depart from the spirit and scope of the present invention are covered within the protection scope of the present invention patent. No reference numerals in the claims should be construed as limiting the scope of the claims. Any simple variations or equivalent substitutions of technical solutions that can be readily obtained by those skilled in the art within the scope of the technology disclosed in the present invention are within the protection scope of the present invention.

Claims

1. A post-earthquake unmanned vehicle-machine collaborative scheduling method based on edge computing, characterized in that, Includes the following steps: Initialize the unmanned vehicle and drone cluster to obtain status data of relief supplies and casualty data; A comprehensive objective function is constructed with the goals of minimizing total rescue time, maximizing safety, minimizing total energy consumption, ensuring fairness in patient transfer, and minimizing communication latency. Specifically, the comprehensive objective function is as follows: ； in, This is the maximum time required to complete the rescue operation. For relief supplies At the receiving point Emergency weight; This refers to positive and negative deviations in material allocation. For receiving point relief supplies Demand; Type of wounded From the receiving point Average arrival time to the rescue station; It is the time threshold for assessing the urgency of the injured. yes Priority weights for different types of casualties; These represent the total energy consumption of the driverless car and the drone, respectively. This represents the total energy consumption and total communication latency of edge computing; The comprehensive objective function is subject to the following constraints: Specifically, to minimize the total rescue time: when material or casualty handover occurs, the time difference between the arrival of unmanned vehicles (UAVs) and drones at the transfer point must be within a set value; the time for an UAV or drone to leave the rescue station must not be less than the sum of its arrival time and the time for loading rescue materials; the time for an UAV or drone to leave the receiving point must not be less than the sum of its arrival time, the time for unloading rescue materials, and the time for picking up casualties, and loading casualties can only begin after the unloading of materials is completed; a final time is set for all vehicles to complete their tasks; to maximize safety, the load of UAVs and drones is set to not exceed the capacity weight constraint; the maximum battery power of drones limits their maximum range; the total data transmission of a single communication link... The transmission capacity must not exceed the bandwidth; to minimize total energy consumption, the energy consumption of unmanned vehicles, drones, and edge computing must not exceed the set limits; the maximum battery capacity of drones limits their maximum range; to ensure fairness in the transfer of wounded personnel, it is stipulated that particularly urgent wounded personnel must be directly transported by drones, and the transfer of extremely urgent wounded personnel is prohibited. Generally, wounded personnel are handed over to unmanned vehicles by drones at the transfer point or transported from the receiving point to the rescue station by unmanned vehicles; spatiotemporal synchronization constraints exist between unmanned vehicles and drones; to minimize communication latency, each computing task must be assigned to an edge server for processing when it is generated; the total latency consisting of transmission latency and computing latency must not exceed the maximum allowable latency; the total computing load of a single edge server must not exceed the computing power limit of the edge server in real time. The comprehensive objective function is solved based on the status data of the relief supplies and the data of the wounded to obtain the scheduling plan for the supplies and the wounded. In the event of aftershocks, roadside units with edge servers are used as edge nodes. Real-time road condition changes, vehicle status, fluctuations in material demand, and new casualty data are obtained based on these edge nodes. The real-time road condition changes, vehicle status, fluctuations in material demand, and new casualty data are fed back to the control center through the edge servers for re-planning and recalculation of unloading, and redistribution of material and casualty dispatch.

2. The post-earthquake unmanned vehicle-machine collaborative scheduling method based on edge computing according to claim 1, characterized in that, The comprehensive objective function is solved based on the status data of the relief supplies and the data of the wounded to obtain the scheduling plan for the supplies and the wounded. Specifically, the Benders decomposition method is used to solve the problem.

3. The post-earthquake unmanned vehicle-machine collaborative scheduling method based on edge computing according to claim 2, characterized in that, The Benders decomposition method decomposes the problem into a main problem and sub-problems. The main problem is resource allocation and casualty evacuation plan, and the sub-problems are path planning and edge computing offloading. The main linear programming problem is solved using CPLEX, while the mixed integer programming subproblem is solved using Gurobi, resulting in a resource and casualty scheduling scheme.

4. The post-earthquake unmanned vehicle-machine collaborative scheduling method based on edge computing according to claim 1, characterized in that, The unmanned vehicles communicate with each other via V2V communication technology to achieve information exchange between vehicles.

5. A post-earthquake unmanned vehicle-machine collaborative scheduling system based on edge computing, characterized in that, include: The initialization module is used to initialize the unmanned vehicle and drone clusters and acquire data on the status of relief supplies and casualties. A comprehensive objective function is constructed with the goals of minimizing total rescue time, maximizing safety, minimizing total energy consumption, ensuring fairness in patient transfer, and minimizing communication latency. Specifically, the comprehensive objective function is as follows: ； in, This is the maximum time required to complete the rescue operation. For relief supplies At the receiving point Emergency weight; This refers to positive and negative deviations in material allocation. For receiving point relief supplies Demand; Type of wounded From the receiving point Average arrival time to the rescue station; It is the time threshold for assessing the urgency of the injured. yes Priority weights for different types of casualties; These represent the total energy consumption of the driverless car and the drone, respectively. This represents the total energy consumption and total communication latency of edge computing; The comprehensive objective function is subject to the following constraints: Specifically, to minimize the total rescue time: when material or casualty handover occurs, the time difference between the arrival of unmanned vehicles (UAVs) and drones at the transfer point must be within a set value; the time for an UAV or drone to leave the rescue station must not be less than the sum of its arrival time and the time for loading rescue materials; the time for an UAV or drone to leave the receiving point must not be less than the sum of its arrival time, the time for unloading rescue materials, and the time for picking up casualties, and loading casualties can only begin after the unloading of materials is completed; a final time is set for all vehicles to complete their tasks; to maximize safety, the load of UAVs and drones is set to not exceed the capacity weight constraint; the maximum battery power of drones limits their maximum range; the total data transmission of a single communication link... The transmission capacity must not exceed the bandwidth; to minimize total energy consumption, the energy consumption of unmanned vehicles, drones, and edge computing must not exceed the set limits; the maximum battery capacity of drones limits their maximum range; to ensure fairness in the transfer of wounded personnel, it is stipulated that particularly urgent wounded personnel must be directly transported by drones, and the transfer of extremely urgent wounded personnel is prohibited. Generally, wounded personnel are handed over to unmanned vehicles by drones at the transfer point or transported from the receiving point to the rescue station by unmanned vehicles; spatiotemporal synchronization constraints exist between unmanned vehicles and drones; to minimize communication latency, each computing task must be assigned to an edge server for processing when it is generated; the total latency consisting of transmission latency and computing latency must not exceed the maximum allowable latency; the total computing load of a single edge server must not exceed the computing power limit of the edge server in real time. The objective optimization module solves the comprehensive objective function based on the status data of the relief supplies and the data of the wounded to obtain a scheduling plan for the supplies and the wounded. The dynamic scheduling module is used to, in the event of aftershocks, use roadside units with edge servers as edge nodes to acquire real-time data on road conditions, vehicle status, fluctuations in material demand, and new casualties. The real-time data on road conditions, vehicle status, fluctuations in material demand, and new casualties is fed back to the control center through the edge servers for re-planning and recalculation of routes and unloading, and for redistribution of material and casualty scheduling.

6. A computer device, comprising a memory, a processor, and a computer program stored in the memory, characterized in that, The processor executes the computer program to implement the steps of the method according to any one of claims 1 to 4.

7. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is loaded by the processor, it is able to perform the steps of the method according to any one of claims 1 to 4.

Images