An electric-hydrogen hybrid unmanned aerial vehicle group task-energy coupling scheduling method and device
By constructing a dual-energy model and an improved non-dominated sorting dung beetle optimization algorithm, combined with reinforcement learning and fuzzy logic, the problems of insufficient energy perception and single scheduling in the UAV swarm system are solved, realizing task-energy collaborative scheduling of electric-hydrogen hybrid UAV swarms, and improving task continuity and swarm collaboration.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- FOSHAN UNIVERSITY
- Filing Date
- 2026-01-22
- Publication Date
- 2026-06-02
Smart Images
Figure CN122131815A_ABST
Abstract
Description
Technical Field
[0001] This invention relates to the field of drone swarm collaborative scheduling and energy management technology, and in particular to a task-energy coupling scheduling method and device for electric-hydrogen hybrid drone swarms, which is applicable to low-altitude drone swarm operation scenarios such as urban inspection, farmland plant protection, and power line inspection. Background Technology
[0002] With the widespread application of drones in energy, power, agriculture, and emergency rescue, drone swarm operations are gradually evolving from single-drone operations to clustered and intelligent operations. In large-area inspections and continuous operation tasks, ensuring the continuous flight capability and collaborative work efficiency of drone swarms has become a core issue for improving system performance.
[0003] Currently, the most common methods for refueling drones are battery charging or replacement, while hydrogen fuel cell drones are gradually being introduced in some scenarios to improve flight time. However, in scenarios involving a hybrid formation of batteries and hydrogen fuel cells, the following prominent issues exist: (1) Insufficient energy perception: Existing group control systems are mainly task-driven and lack dynamic perception and modeling of the remaining energy and recharge rhythm of UAVs of different energy types.
[0004] (2) Single scheduling mechanism: Existing methods focus on task allocation and path planning, making it difficult to adjust the task priority and operation sequence of UAVs in real time according to energy status.
[0005] (3) Energy-task separation: Electric and hydrogen-powered UAVs are not coordinated enough in the same mission system and lack a complete energy-task joint optimization scheduling strategy, which leads to some UAVs returning to base too early or having energy redundancy.
[0006] (4) Limited efficiency: Due to the lack of an energy-driven scheduling mechanism, the existing system is unable to fully leverage the group operation advantages of electric-hydrogen hybrid fleets in complex scenarios with intersecting paths and overlapping tasks.
[0007] Therefore, there is an urgent need to propose a joint optimization method that can simultaneously consider the UAV's energy type, remaining energy, recharge cycle, and task scheduling logic in order to break through the technical bottleneck of existing group control scheduling. Summary of the Invention
[0008] The purpose of this invention is to propose a task-energy coupling scheduling method and device for electric-hydrogen hybrid unmanned aerial vehicle (UAV) swarms. By constructing a dual-energy model, the method optimizes task continuity in a multi-UAV collaborative state. The specific technical solution is as follows: A task-energy coupling scheduling method for hybrid electric-hydrogen unmanned aerial vehicle (UAV) swarms includes the following steps: Obtain the energy type, remaining energy, and refueling cycle of each drone in the formation; Set task priorities and time windows based on task type; The remaining energy, replenishment cycle, and operational capacity are included in the weighting calculation to obtain the weighting factors; Construct a joint task-energy objective optimization function, and solve for the optimal scheduling scheme of the UAV swarm based on the joint task-energy objective optimization function; The flight paths and task execution order of each UAV are dynamically updated based on the optimal scheduling scheme of the UAV swarm, ensuring that the overall task completion rate and energy utilization of the swarm are optimal.
[0009] The proposed task-energy coupling scheduling method for electric-hydrogen hybrid UAV swarms enables task-energy coordinated scheduling of UAV swarms under electric-hydrogen hybrid energy mode, improving task continuity and swarm coordination, and providing support for UAV swarm operations in future complex task environments.
[0010] Preferably, the electric-hydrogen hybrid unmanned aerial vehicle swarm mission-energy coupling scheduling method further includes the following steps: Real-time monitoring of energy and mission progress; When there is insufficient energy or task delays, the adaptive control logic is triggered to readjust task allocation and paths.
[0011] Preferably, the specific method for constructing the joint task-energy objective optimization function includes the following steps: Obtain the delay, the corresponding revenue weight, and the penalty for non-completion for each task. Construct a first objective sub-function to minimize the time default and non-completion penalty based on the delay, service duration, and penalty for non-completion. Construct a joint task-energy objective optimization function based on the first objective sub-function.
[0012] Preferably, the specific method for constructing the joint objective optimization function of task and energy further includes the following steps: The system obtains the energy consumption per unit distance, the flight distance of the UAV, the energy consumption per unit time of task execution, the service duration, and the station recharging duration. Based on these parameters, a second objective subfunction is obtained to minimize the total lifecycle operation energy consumption. The task-energy joint objective optimization function is constructed based on the first objective sub-function as follows: The task-energy joint objective optimization function is constructed based on the first objective sub-function and the second objective sub-function.
[0013] Preferably, the specific method for constructing the joint objective optimization function of task and energy further includes the following steps: Obtain the remaining energy of the drone before it arrives at the mission and the level of site congestion. Based on the remaining energy and the level of site congestion, obtain a third objective sub-function to minimize energy risk and site congestion. The task-energy joint objective optimization function is constructed based on the first objective sub-function and the second objective sub-function. Specifically, the task-energy joint objective optimization function is constructed based on the first objective sub-function, the second objective sub-function, and the third objective sub-function.
[0014] Preferably, the electric-hydrogen hybrid unmanned aerial vehicle swarm mission-energy coupling scheduling method further includes the following steps: Obtain the constraints of unique task allocation and execution consistency, path continuity and flow conservation, time window and timing constraints, energy dynamic constraints, energy replenishment process constraints, and site capacity and congestion constraints; Constraints are constructed based on the constraints of unique task allocation and execution consistency, path continuity and flow conservation, time window and timing constraints, energy dynamic constraints, energy replenishment process constraints, and site capacity and congestion constraints.
[0015] A task-energy coupling scheduling device for an electric-hydrogen hybrid unmanned aerial vehicle (UAV) swarm, used to implement the aforementioned task-energy coupling scheduling method for an electric-hydrogen hybrid UAV swarm, comprising: The energy status perception module is used to obtain the energy type, remaining energy, and recharging cycle of each drone in the formation; The task requirements modeling module is used to set task priorities and time windows based on task type; The weighting factor calculation module is used to incorporate remaining energy, replenishment cycle and operating capacity into the weighting calculation to obtain the weighting factor; The coupled optimization scheduling module is used to construct the task-energy joint objective optimization function, and based on the task-energy joint objective optimization function, solve the optimal scheduling scheme of the UAV swarm; The path order adjustment module is used to dynamically update the flight path and task execution order of each UAV according to the optimal scheduling scheme of the UAV swarm, so as to ensure the optimal overall task completion rate and energy utilization of the swarm.
[0016] Preferably, the electric-hydrogen hybrid unmanned aerial vehicle swarm mission-energy coupling scheduling device further includes: The scheduling platform is used to monitor energy and task progress in real time. When there is insufficient energy or task delays, it triggers adaptive control logic to readjust task allocation and paths.
[0017] Preferably, the coupling optimization scheduling module includes: The first objective sub-function construction unit is used to obtain the delay, the corresponding revenue weight, and the penalty for non-completion for each task, and to construct a first objective sub-function to minimize the time default and non-completion penalty based on the delay, service duration, and non-completion penalty. The second objective sub-function construction unit is used to obtain the energy consumption per unit distance, the UAV's flight distance, the energy consumption per unit time for task execution, the service duration, and the site refueling duration. Based on the energy consumption per unit distance, the UAV's flight distance, the energy consumption per unit time for task execution, the service duration, and the site refueling duration, a second objective sub-function is obtained to minimize the total lifecycle operation energy consumption. The third objective sub-function construction unit is used to obtain the remaining energy of the UAV before it arrives at the mission and the site congestion level. Based on the remaining energy and the site congestion level, a third objective sub-function is obtained to minimize energy risk and site congestion. The joint objective optimization function construction unit constructs the task-energy joint objective optimization function based on the first objective sub-function and the second objective sub-function. Specifically, it constructs the task-energy joint objective optimization function based on the first objective sub-function, the second objective sub-function, and the third objective sub-function.
[0018] Preferably, the electric-hydrogen hybrid unmanned aerial vehicle swarm mission-energy coupling scheduling device further includes: The constraint construction module is used to obtain constraints on task unique allocation and execution consistency, path continuity and flow conservation, time window and timing constraints, energy dynamic constraints, energy replenishment process constraints, and site capacity and congestion constraints, and to construct constraint conditions based on these constraints. Attached Figure Description
[0019] The invention will be further understood from the following description taken in conjunction with the accompanying drawings. The components in the drawings are not necessarily drawn to scale, but rather the emphasis is on illustrating the principles of the embodiments. In different views, the same reference numerals designate corresponding parts.
[0020] Figure 1 This is a schematic diagram of the overall process of a task-energy coupling scheduling method for an electric-hydrogen hybrid unmanned aerial vehicle swarm according to an embodiment of the present invention; Figure 2 This is a flowchart illustrating a specific method for constructing a joint task-energy objective optimization function in one embodiment of the present invention; Figure 3 This is a flowchart illustrating a task-energy coupling scheduling method for an electric-hydrogen hybrid unmanned aerial vehicle swarm, as described in another embodiment of the present invention. Figure 1 ; Figure 4This is a flowchart illustrating a task-energy coupling scheduling method for an electric-hydrogen hybrid unmanned aerial vehicle swarm, as described in another embodiment of the present invention. Figure 2 ; Figure 5 This is a schematic diagram of the improved non-dominated sorting dung beetle optimization algorithm in one embodiment of the present invention; Figure 6 This is a flowchart of the dynamic task scheduling process for unmanned aerial vehicles (UAVs) in one embodiment of the present invention. Detailed Implementation
[0021] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be further described in detail below with reference to its embodiments. It should be understood that the specific embodiments described herein are merely illustrative of the invention and do not limit the scope of protection of the invention.
[0022] It should be noted that when an element is referred to as being "fixed to" another element, it can be directly attached to the other element or there may be an intervening element. When an element is referred to as being "connected to" another element, it can be directly connected to the other element or there may be an intervening element. The terms "upper," "lower," "left," "right," and similar expressions used herein are for illustrative purposes only and do not represent the only possible implementation.
[0023] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention pertains. The terminology used herein in the description of this invention is for the purpose of describing particular embodiments only and is not intended to be limiting of the invention. The terms "and" and "or" as used herein include any and all combinations of one or more of the associated listed items.
[0024] like Figure 1 As shown, an embodiment of the present invention provides a task-energy coupling scheduling method for an electric-hydrogen hybrid unmanned aerial vehicle swarm, comprising the following steps: S1: Obtain the energy type, remaining energy, and refueling cycle of each drone in the formation.
[0025] Specifically, the energy types include batteries / hydrogen fuel cells.
[0026] S2 sets task priority and time window based on task type.
[0027] The types of tasks include inspection, plant protection, and power transmission channel inspection.
[0028] S3 incorporates remaining energy, recharge cycle, and operational capacity into the weight calculation to obtain the weighting factor.
[0029] S4. Construct a joint task-energy objective optimization function, and based on the joint task-energy objective optimization function, solve for the optimal scheduling scheme of the UAV swarm.
[0030] The specific method for constructing the task-energy joint objective optimization function includes the following steps: obtaining the delay of each task, the corresponding revenue weight, and the penalty for non-completion; constructing a first objective sub-function to minimize time default and non-completion penalty based on the delay, service duration, and non-completion penalty; and constructing the task-energy joint objective optimization function based on the first objective sub-function.
[0031] The specific method for constructing the task-energy joint objective optimization function also includes the following steps: obtaining the energy consumption per unit distance, the UAV's flight distance, the energy consumption per unit time for task execution, the service duration, and the site refueling duration; and obtaining a second objective sub-function to minimize the total lifecycle operation energy consumption based on the energy consumption per unit distance, the UAV's flight distance, the energy consumption per unit time for task execution, the service duration, and the site refueling duration. Specifically, constructing the task-energy joint objective optimization function based on the first objective sub-function involves: constructing the task-energy joint objective optimization function based on the first objective sub-function and the second objective sub-function.
[0032] The specific method for constructing the joint task-energy objective optimization function also includes the following steps: obtaining the remaining energy of the UAV before arriving at the mission and the site congestion level, and obtaining a third objective sub-function to minimize energy risk and site congestion based on the remaining energy and site congestion level. Specifically, constructing the joint task-energy objective optimization function based on the first and second objective sub-functions involves: constructing the joint task-energy objective optimization function based on the first, second, and third objective sub-functions.
[0033] As a preferred technical solution, such as Figure 2 As shown, the specific method for constructing the joint task-energy objective optimization function includes the following steps: S41, obtain the delay of each task, the corresponding revenue weight, and the penalty for non-completion. Based on the delay, service duration, and penalty for non-completion, construct a first objective sub-function to minimize the time default and penalty for non-completion.
[0034] S42, obtain the energy consumption per unit distance, the flight distance of the UAV, the energy consumption per unit time of task execution, the service duration, and the station recharging duration, and obtain the second objective sub-function for minimizing the energy consumption of the entire life cycle operation based on the energy consumption per unit distance, the flight distance of the UAV, the energy consumption per unit time of task execution, the service duration, and the station recharging duration.
[0035] S43: Obtain the remaining energy of the drone before it arrives at the mission and the site congestion level. Based on the remaining energy and the site congestion level, obtain a third objective sub-function to minimize energy risk and site congestion.
[0036] S44. Construct a joint objective optimization function for the task and energy based on the first objective sub-function, the second objective sub-function, and the third objective sub-function.
[0037] Specifically, based on the dual-energy model, a joint objective optimization function of task and energy can be constructed, and the optimal scheduling scheme of the UAV swarm can be solved by using the improved non-dominated Sorting Dung Beetle Optimizer (INSDBO) algorithm.
[0038] The first objective function is expressed as follows: Among them, weight This indicates the importance or priority of a task; for example, higher priority tasks have greater weight. Used to penalize delayed tasks, weight The higher the value, the heavier the penalty. 'i' represents the task index, indicating a single task within set I. This represents the penalty coefficient for not completing a task, used to quantify the degree of punishment for not executing a task. The larger the value, the more the system tends to execute all tasks.
[0039] Because drone swarm tasks (such as inspection and plant protection) have significantly different priorities, using static weights can easily lead to inefficient scheduling under unexpected events (such as weather changes or emergency tasks), and it cannot adapt to complex scenarios. For example, defaults on high-priority tasks have a greater impact. As a preferred technical solution, a dynamic adaptive penalty function based on reinforcement learning and fuzzy logic is introduced to incorporate static penalty coefficients (such as weights) into the overall performance. and / or penalty coefficient for not completing the task The system is upgraded to an adaptive dynamic weighting mechanism, integrating reinforcement learning and fuzzy logic. Specifically, a reinforcement learning model, such as a Q-learning model, is first built and trained. The weight coefficients are dynamically adjusted based on real-time scenarios (such as task priority, remaining energy, and historical default data) to reward timely task completion. Then, a fuzzy logic module is built to handle uncertainties such as "low energy risk" or "high task urgency" using fuzzy rules, transforming qualitative descriptions into quantitative weights and enhancing the robustness of the function to nonlinear scenarios.
[0040] More specifically, weight =RL(Priority level of task i, remaining energy status of drone u, historical scheduling records) + Fuzzy(Real-time risk level of task i). For task i's priority level, a higher priority level results in a greater impact from default, such as power line inspection > farmland protection. Historical scheduling records include the number of defaults and their causes (insufficient energy / path congestion). The real-time risk level of task i can be determined based on imprecise information such as weather warnings, sudden obstacles, and the probability of energy station congestion.
[0041] For the reinforcement learning (RL) model, the current task priority, the UAV's energy state, and recent historical data slices serve as the state space, while the output's basic weight values form the action space. Instantaneous rewards are calculated based on task completion rate and energy utilization rate. For the fuzzy logic module, the real-time risk level of task i is used as input, and preset fuzzy rules such as "IF task i's real-time risk level IS High THEN Δ" are applied. =+0.2” or “IF task priority is high AND energy risk is high THEN Δ” =+0.3”, etc., to obtain the weight. Adjustment amount Δ This can be converted into an accurate value through defuzzification. Similarly, the penalty coefficient for incomplete tasks can also be dynamically adjusted based on the reinforcement learning model and fuzzy logic module.
[0042] Ultimately, the dynamically adjusted weight coefficients are equal to the base weight values output by the reinforcement learning model (RL) plus the weight adjustment amount output by the fuzzy logic module (Fuzzy).
[0043] In this way, the problem of low scheduling efficiency and inability to adapt to complex scenarios caused by static weights can be solved by using a dynamic weight allocation mechanism.
[0044] Here, the first objective sub-function aims to reduce the penalties for task delays and unexecuted tasks, ensuring timely task completion. This is achieved through weighting... In addition, the penalty coefficient for unfinished tasks is dynamically adjusted to adjust task allocation, ensuring that high-priority tasks are completed first and reducing omissions.
[0045] The second objective function is expressed as follows: This is used to minimize the energy consumption of the entire lifecycle operation (including flight, operation, and refueling losses). Here, u represents the UAV index, which represents a single UAV in the set U, A represents the flight path between network nodes (such as task points and refueling stations), and (i,j) is the arc index, which represents the flight path from node i to node j. The energy type (electric or hydrogen-powered) of the drone is used to distinguish its energy consumption characteristics. This represents the total energy consumption of the drone during a flight segment. Service duration. Indicates the time required to perform the task. This indicates the total energy consumption for task execution. The refueling loss coefficient for UAV u represents the energy loss rate during the refueling process and depends on the energy type value. The smaller the value, the higher the efficiency. This represents the total energy lost during energy replenishment.
[0046] Here, the second objective function aims to optimize the energy consumption of the UAV throughout the entire process from flight and mission execution to refueling, covering both electric and hydrogen-powered energy types. It models the differences between electric and hydrogen-powered UAVs by distinguishing energy types and loss coefficients, so as to minimize the overall energy consumption from flight to refueling and improve energy utilization.
[0047] The third objective function is expressed as: Among them, . This represents the energy risk penalty coefficient, used to quantify the severity of energy shortages. A higher value indicates a greater tendency for the system to maintain a high surplus of energy. Safe energy threshold. This indicates the minimum remaining energy that a drone should maintain to avoid risks. Indicates the safety energy ratio coefficient. This represents the station congestion penalty coefficient, used to quantify the severity of congestion. The larger the value, the more the system tends to distribute power supply. The congestion level of station s can be approximated by the peak occupancy of that station. This indicates congestion penalties.
[0048] Here, the third objective subfunction aims to reduce the risk of drone energy depletion and refueling station congestion, ensuring operational safety and efficiency. It achieves dynamic monitoring through a safe energy threshold and congestion penalties.
[0049] As a preferred technical solution, such as Figure 3 As shown, the electric-hydrogen hybrid unmanned aerial vehicle swarm mission-energy coupling scheduling method further includes the following steps: S45, obtain the unique allocation and execution consistency constraints of tasks, path continuity and flow conservation constraints, time window and timing constraints, energy dynamic constraints, energy replenishment process constraints, and site capacity and congestion constraints.
[0050] S46. Constraints are constructed based on the unique allocation and execution consistency constraints of tasks, path continuity and flow conservation constraints, time window and timing constraints, energy dynamic constraints, energy replenishment process constraints, and site capacity and congestion constraints.
[0051] For example, the constraint on unique task allocation and execution consistency is expressed as follows: Specifically, for each task i, there must be one and only one drone u to prevent tasks from being duplicated or omitted, and to ensure the uniqueness of task execution.
[0052] The path continuity and flow conservation constraints are expressed as: 1) Take-off and landing point 0: This means that each drone u must start from the take-off and landing point (node 0) and go to a certain node j, ensuring that each drone starts its path from the take-off and landing point.
[0053] 2) General nodes (task and power replenishment nodes): This means that for each non-take-off and landing node i (such as a mission point or refueling station), the number of times the drone u enters the node must be equal to the number of times it leaves the node, in order to maintain the flow conservation of the path and avoid the drone being stuck or interrupted at the node.
[0054] 3) Binding of task access and execution: This means that if drone u performs task i, it must access the task node, and vice versa. By binding task execution decisions and path access decisions, consistency between the two is ensured.
[0055] Time windows and timing constraints are expressed as follows: 1) Arrival time of arc triggering: This means that if drone u flies from node i to node j, the arrival time j must be greater than or equal to the arrival time at i plus the service time and flight time. In other words, the time for the drone to arrive at node j must take into account the end of the service at the previous node and the flight time to ensure the time continuity of the path.
[0056] 2) Task Time Window: This means that the time when the drone u arrives at task i must be within the task's time window, so as to force the task to be executed within the specified time window and prevent it from starting too early or too late.
[0057] 3) Task completion time aggregation: This means that the completion time of task i is equal to the arrival time plus the service time.
[0058] The energy dynamic constraint is expressed as: 1) Initial energy: , indicates that the remaining energy of the drone u at takeoff is equal to its initial energy, which is not greater than the maximum possible energy of the drone body. It is mainly used to set the initial energy state of the drone.
[0059] 2) Arc energy transfer: This indicates that if the drone u flies from node i to node j, the remaining energy is used at j to reduce flight energy consumption (related to distance) and mission execution energy consumption. It is mainly used to simulate the dynamic consumption of energy during flight and mission execution.
[0060] Lower limit of safe energy: This means that the remaining energy of the drone u when it arrives at any node i must not be lower than the safe energy threshold, in order to prevent the risk of energy depletion and ensure that the drone has enough energy to return or recharge.
[0061] Upper energy realm: This means that the remaining energy of the drone u when it arrives at any node i cannot exceed the maximum available energy, in order to prevent energy overload.
[0062] The energy replenishment process constraint (distinguishing between electrical energy and hydrogen energy) is expressed as: 1) Energy replenishment duration: Of which, electrical energy is Hydrogen energy is .
[0063] 2) Power restoration: It represents the energy of the electric drone after recharging at station s. equal to the initial energy In addition, effective energy restoration; among them, .
[0064] Generally, refueling time depends on the energy type. For electric drones, refueling time is based on charging power and efficiency, while for hydrogen-powered drones it's based on refueling rate and efficiency. The effective restored energy of an electric drone equals the product of charging power, efficiency, and refueling time; the effective restored energy of a hydrogen-powered drone equals the product of refueling rate, efficiency, and refueling time.
[0065] 3) Energy replenishment timing binding: If site s is accessed, the time / energy variables take effect, which can be triggered by an arc. Binding , Arrival / departure time; or consider station replenishment as a "service node" with service duration.
[0066] The site capacity and congestion constraints are expressed as follows: 1) No time limit for concurrency (The capacity of site s represents the maximum number of drones that can be served simultaneously): Modeled using discrete time slices or non-overlapping intervals.
[0067] The discretization formula is as follows: This indicates that at any time slice t, the total number of drones replenishing power at station s does not exceed its capacity, ensuring that the station does not operate under overload conditions; where time slice... , For the length of the film, express exist interval Energy is being replenished.
[0068] The table below shows the definitions of some variable sets and related parameters involved in the mathematical modeling based on MILP (Mixed Integer Linear Programming) in this invention.
[0069]
[0070] S5 dynamically updates the flight paths and task execution order of each UAV based on the solved optimal scheduling scheme for the UAV swarm, ensuring optimal overall task completion and energy utilization for the swarm.
[0071] In summary, the proposed task-energy coupling scheduling method for electric-hydrogen hybrid UAV swarms enables task-energy coordinated scheduling of UAV swarms under electric-hydrogen hybrid energy mode, improving task continuity and swarm coordination, and providing support for UAV swarm operations in complex task environments in the future.
[0072] In one embodiment, such as Figure 4 As shown, the electric-hydrogen hybrid unmanned aerial vehicle swarm mission-energy coupling scheduling method further includes the following steps: S6 provides real-time monitoring of energy and mission progress.
[0073] S7 triggers adaptive control logic to readjust task allocation and paths when there is insufficient energy or task delays.
[0074] In this way, by monitoring energy and task progress in real time, and triggering adaptive control logic when there is insufficient energy or task delays, task allocation and paths can be readjusted, thus achieving adaptive control and feedback.
[0075] The specific method for solving the optimal scheduling scheme of UAV swarm based on the joint objective optimization function of task and energy includes: solving the joint objective optimization function of task and energy based on the improved non-dominated sorting dung beetle optimization algorithm to obtain the optimal scheduling scheme of UAV swarm.
[0076] like Figure 5 as well as Figure 6 As shown, the improved non-dominated sorting dung beetle optimization algorithm for solving the joint task-energy objective function includes the following steps: 1. Encoding and Decoding Encoding method: A chromosome is used to represent a set of drone task assignments and paths.
[0077] Decoding rules: Simulate the execution process of the drone in sequence; calculate flight time, mission completion time, and energy consumption; if energy is insufficient, insert a refueling station and allocate refueling time; if the mission time window is exceeded, add a late penalty.
[0078] 2. Fitness function calculation In the UAV swarm task-energy coupled scheduling problem, each individual (i.e., a scheduling scheme) can be decoded to obtain the task execution order, UAV path, and energy replenishment strategy. Fitness function calculation consists of two parts: objective function value calculation and constraint penalty handling. The task completion time and remaining energy are obtained by decoding the path and energy replenishment strategy; then, the objective function value is calculated... It also checks constraints and calculates penalty terms; finally, it synthesizes results based on fitness. Input a non-dominated sort to determine the quality of individuals.
[0079] 3. Initialize the population Since the initial population of the dung beetle optimization algorithm is random, it cannot guarantee a uniform distribution of the initial positions of individuals in the search space, affecting the algorithm's search speed and optimization performance. The improved non-dominated sorting dung beetle algorithm introduces a Piecewise chaotic mapping to generate the initial population during the initialization process. This increases the ergodicity of the initial population, making the population distribution more even. Each individual is decoded, its fitness is calculated, a non-dominated sort is performed, and the initial Pareto front solutions are selected and stored in the external archive.
[0080] 4. A random walk strategy was used to perturb the dung beetle's ball-rolling behavior. In the original dung beetle optimization algorithm, the rolling ball tends to move along inertia or prior directions, easily getting trapped in local optima. To improve this limitation, a random walk strategy is introduced. This strategy effectively enhances the algorithm's ability to escape local optima through refined local exploration and adaptive perturbation mechanisms, thereby improving the balance between global exploration and local fine-tuning. To maintain algorithm efficiency, this algorithm triggers the random walk strategy if the optimal value has not been updated within 5 iterations, thus avoiding redundant calculations and maintaining low computational complexity. The random walk process is as follows: (18) In the formula: The set of steps for a random walk; This is the cumulative sum of steps taken. This represents the current number of steps taken. For the first Steps taken per trip.
[0081] 5. Evolutionary Iteration Perform the following steps in each iteration: (1) Dung beetle position update (DBO operator): Guided update: Individuals move closer to elite solutions (in Archive); Local perturbation: Randomly swap task order, adjust energy replenishment sites or duration; Group learning: Learn from the current best individual to increase search efficiency.
[0082] (2) Decoding and repair: Check if the path is feasible (time window, energy); if not, insert a replenishment station / adjust the execution order to repair.
[0083] (3) Non-dominated sorting and population selection: merge the parent generation and the offspring generation to obtain the candidate set. Perform non-dominated sorting to obtain several frontier layers; use crowding distance screening to maintain a fixed population size.
[0084] (4) Update external archives: merge newly generated non-dominated solutions into Archive; if Archive is overloaded, delete solutions with low crowding to maintain diversity.
[0085] 6. Rolling time-domain monitoring (dynamic scheduling mechanism) Scroll optimization is triggered if any of the following conditions are met: 1) The SoE (State of Energy) of a certain drone is approaching the safe threshold. .
[0086] 2) A task is expected to time out (e.g.) ) 3) Congestion occurs at a certain power replenishment site (e.g., the number of drones in queue exceeds the capacity). ) Rolling optimization process: Reinitialize the population, using the current state of the drone swarm as input to regenerate the initial population; use the previous scheduling result as part of the population, which is equivalent to "hot start" and accelerates convergence; re-execute the evaluation, non-dominated sorting, position update, decoding and repair steps to obtain a new Pareto solution set; select a new optimal scheduling scheme from it.
[0087] 6. Results Output and Scheduling Scheme Selection If the maximum number of iterations (MaxGen) is reached or the convergence criterion (Pareto front solution is no longer improved) is met, and rolling optimization is no longer triggered, the iteration is terminated. Output the Pareto front solution set in Archive; select a scheduling scheme based on user preference (e.g., "time-first" or "energy-first"); and distribute the selected scheme to the drone swarm for execution.
[0088] An embodiment of the present invention also provides a task-energy coupling scheduling device for an electric-hydrogen hybrid unmanned aerial vehicle (UAV) swarm, used to implement the aforementioned task-energy coupling scheduling method for an electric-hydrogen hybrid UAV swarm, comprising an energy status perception module, a task requirement modeling module, a weight factor calculation module, a coupling optimization scheduling module, and a path order adjustment module.
[0089] The energy status perception module is used to obtain the energy type, remaining energy, and refueling cycle of each UAV in the formation; the task requirement modeling module is used to set task priority and time window according to task type; the weight factor calculation module is used to incorporate remaining energy, refueling cycle, and operational capability into weight calculation to obtain weight factors.
[0090] The coupled optimization scheduling module is used to construct a joint task-energy objective optimization function, and based on the joint task-energy objective optimization function, solve the optimal scheduling scheme for the UAV swarm. The path order adjustment module is used to dynamically update the flight path and task execution order of each UAV according to the solved optimal scheduling scheme for the UAV swarm, so as to ensure the optimal overall task completion and energy utilization of the swarm.
[0091] The hybrid electric unmanned aerial vehicle (UAV) swarm mission-energy coupling scheduling device also includes a scheduling platform. The scheduling platform is used to monitor energy and mission progress in real time. In the event of insufficient energy or mission delays, it triggers adaptive control logic to readjust mission allocation and paths.
[0092] Specifically, the coupled optimization scheduling module includes a first objective sub-function construction unit, a second objective sub-function construction unit, and a third objective sub-function construction unit.
[0093] The first objective sub-function construction unit is used to obtain the delay of each task, the corresponding revenue weight, and the penalty for non-completion. Based on the delay, service duration, and penalty for non-completion, a first objective sub-function is constructed to minimize time default and penalty for non-completion. The second objective sub-function construction unit is used to obtain the energy consumption per unit distance, the drone's flight distance, the energy consumption per unit time for task execution, the service duration, and the site refueling time. Based on the energy consumption per unit distance, the drone's flight distance, the energy consumption per unit time for task execution, the service duration, and the site refueling time, a second objective sub-function is constructed to minimize the energy consumption of the entire life cycle operation. The third objective sub-function construction unit is used to obtain the remaining energy of the drone before arriving at the task and the site congestion level. Based on the remaining energy and the site congestion level, a third objective sub-function is constructed to minimize energy risk and site congestion.
[0094] The joint objective optimization function construction unit constructs the task-energy joint objective optimization function based on the first objective sub-function and the second objective sub-function. Specifically, it constructs the task-energy joint objective optimization function based on the first objective sub-function, the second objective sub-function, and the third objective sub-function.
[0095] The electric-hydrogen hybrid unmanned aerial vehicle swarm mission-energy coupling scheduling device also includes a constraint construction module.
[0096] The constraint construction module is used to obtain constraints on unique task allocation and execution consistency, path continuity and flow conservation, time window and timing constraints, energy dynamic constraints, energy replenishment process constraints, and site capacity and congestion constraints. Constraint conditions are constructed based on these constraints.
[0097] The electric-hydrogen hybrid UAV swarm task-energy coupling scheduling device described in this invention can realize task-energy coordinated scheduling of UAV swarms in electric-hydrogen hybrid energy mode, improve task continuity and group coordination, and provide support for UAV swarm operations in complex task environments in the future.
[0098] In summary, the present invention has the following advantages: (1) Energy type modeling and task allocation weighting mechanism: For electric drones and hydrogen-powered drones, task allocation weighting factors are established respectively, and the task allocation strategy is dynamically adjusted according to the differences in energy characteristics.
[0099] (2) Task-energy coupling optimization model: Construct a joint optimization scheduling model that takes into account the remaining energy, replenishment cycle and operation capacity to ensure that the overall operation time of the group is maximized.
[0100] (3) Adaptive adjustment of path and execution order: The path planning and task execution order of the UAV are dynamically updated based on the scheduling results to achieve the optimal completion of the task under energy constraints.
[0101] (4) Adaptive control logic under multiple constraints: The scheduling platform has the ability to adapt and adjust under multiple constraints such as energy type, task time, and refueling window to improve the overall operational efficiency of the UAV swarm.
[0102] The technical features of the above embodiments can be combined in any way. For the sake of brevity, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.
[0103] The embodiments described above are merely illustrative of several implementations of the present invention, and while the descriptions are relatively specific and detailed, they should not be construed as limiting the scope of the invention patent. It should be noted that those skilled in the art can make various modifications and improvements without departing from the concept of the present invention, and these all fall within the protection scope of the present invention. Therefore, the protection scope of this invention patent should be determined by the appended claims.
Claims
1. A task-energy coupling scheduling method for electric-hydrogen hybrid unmanned aerial vehicle swarms, characterized in that, The electric-hydrogen hybrid unmanned aerial vehicle (UAV) swarm mission-energy coupling scheduling method includes the following steps: Obtain the energy type, remaining energy, and refueling cycle of each drone in the formation; Set task priorities and time windows based on task type; The remaining energy, replenishment cycle, and operational capacity are included in the weighting calculation to obtain the weighting factors; Construct a joint task-energy objective optimization function, and solve for the optimal scheduling scheme of the UAV swarm based on the joint task-energy objective optimization function; The flight paths and task execution order of each UAV are dynamically updated based on the optimal scheduling scheme of the UAV swarm, ensuring that the overall task completion rate and energy utilization of the swarm are optimal.
2. The task-energy coupling scheduling method for an electric-hydrogen hybrid unmanned aerial vehicle swarm as described in claim 1, characterized in that, The electric-hydrogen hybrid unmanned aerial vehicle (UAV) swarm mission-energy coupling scheduling method further includes the following steps: Real-time monitoring of energy and mission progress; When there is insufficient energy or task delays, the adaptive control logic is triggered to readjust task allocation and paths.
3. The task-energy coupling scheduling method for an electric-hydrogen hybrid unmanned aerial vehicle swarm as described in claim 2, characterized in that, The specific method for constructing the joint objective function of task and energy includes the following steps: Obtain the delay, the corresponding revenue weight, and the penalty for non-completion for each task. Construct a first objective sub-function to minimize the time default and non-completion penalty based on the delay, service duration, and penalty for non-completion. Construct a joint task-energy objective optimization function based on the first objective sub-function.
4. The task-energy coupling scheduling method for an electric-hydrogen hybrid unmanned aerial vehicle swarm as described in claim 3, characterized in that, The specific method for constructing the joint task-energy objective optimization function also includes the following steps: The system obtains the energy consumption per unit distance, the flight distance of the UAV, the energy consumption per unit time of task execution, the service duration, and the station recharging duration. Based on these parameters, a second objective subfunction is obtained to minimize the total lifecycle operation energy consumption. The task-energy joint objective optimization function is constructed based on the first objective sub-function as follows: The task-energy joint objective optimization function is constructed based on the first objective sub-function and the second objective sub-function.
5. The task-energy coupling scheduling method for an electric-hydrogen hybrid unmanned aerial vehicle swarm as described in claim 4, characterized in that, The specific method for constructing the joint task-energy objective optimization function also includes the following steps: Obtain the remaining energy of the drone before it arrives at the mission and the level of site congestion. Based on the remaining energy and the level of site congestion, obtain a third objective sub-function to minimize energy risk and site congestion. The task-energy joint objective optimization function is constructed based on the first objective sub-function and the second objective sub-function. Specifically, the task-energy joint objective optimization function is constructed based on the first objective sub-function, the second objective sub-function, and the third objective sub-function.
6. The task-energy coupling scheduling method for an electric-hydrogen hybrid unmanned aerial vehicle swarm as described in claim 5, characterized in that, The electric-hydrogen hybrid unmanned aerial vehicle (UAV) swarm mission-energy coupling scheduling method further includes the following steps: Obtain the constraints of unique task allocation and execution consistency, path continuity and flow conservation, time window and timing constraints, energy dynamic constraints, energy replenishment process constraints, and site capacity and congestion constraints; Constraints are constructed based on the constraints of unique task allocation and execution consistency, path continuity and flow conservation, time window and timing constraints, energy dynamic constraints, energy replenishment process constraints, and site capacity and congestion constraints.
7. A task-energy coupling scheduling device for an electric-hydrogen hybrid unmanned aerial vehicle (UAV) swarm, used to implement the task-energy coupling scheduling method for an electric-hydrogen hybrid UAV swarm as described in any one of claims 1-6, characterized in that, The electric-hydrogen hybrid unmanned aerial vehicle swarm mission-energy coupling scheduling device includes: The energy status perception module is used to obtain the energy type, remaining energy, and recharging cycle of each drone in the formation; The task requirements modeling module is used to set task priorities and time windows based on task type; The weighting factor calculation module is used to incorporate remaining energy, replenishment cycle and operating capacity into the weighting calculation to obtain the weighting factor; The coupled optimization scheduling module is used to construct the task-energy joint objective optimization function, and based on the task-energy joint objective optimization function, solve the optimal scheduling scheme of the UAV swarm; The path order adjustment module is used to dynamically update the flight path and task execution order of each UAV according to the optimal scheduling scheme of the UAV swarm, so as to ensure the optimal overall task completion rate and energy utilization of the swarm.
8. The electric-hydrogen hybrid unmanned aerial vehicle swarm mission-energy coupling scheduling device as described in claim 7, characterized in that, The electric-hydrogen hybrid unmanned aerial vehicle swarm mission-energy coupling scheduling device also includes: The scheduling platform is used to monitor energy and task progress in real time. When there is insufficient energy or task delays, it triggers adaptive control logic to readjust task allocation and paths.
9. The electric-hydrogen hybrid unmanned aerial vehicle swarm mission-energy coupling scheduling device as described in claim 8, characterized in that, The coupling optimization scheduling module includes: The first objective sub-function construction unit is used to obtain the delay, the corresponding revenue weight, and the penalty for non-completion for each task, and to construct a first objective sub-function to minimize the time default and non-completion penalty based on the delay, service duration, and non-completion penalty. The second objective sub-function construction unit is used to obtain the energy consumption per unit distance, the UAV's flight distance, the energy consumption per unit time for task execution, the service duration, and the site refueling duration. Based on the energy consumption per unit distance, the UAV's flight distance, the energy consumption per unit time for task execution, the service duration, and the site refueling duration, a second objective sub-function is obtained to minimize the total lifecycle operation energy consumption. The third objective sub-function construction unit is used to obtain the remaining energy of the UAV before it arrives at the mission and the site congestion level. Based on the remaining energy and the site congestion level, a third objective sub-function is obtained to minimize energy risk and site congestion. The joint objective optimization function construction unit constructs the task-energy joint objective optimization function based on the first objective sub-function and the second objective sub-function. Specifically, it constructs the task-energy joint objective optimization function based on the first objective sub-function, the second objective sub-function, and the third objective sub-function.
10. The electric-hydrogen hybrid unmanned aerial vehicle swarm mission-energy coupling scheduling device as described in claim 9, characterized in that, The electric-hydrogen hybrid unmanned aerial vehicle swarm mission-energy coupling scheduling device also includes: The constraint construction module is used to obtain constraints on task unique allocation and execution consistency, path continuity and flow conservation, time window and timing constraints, energy dynamic constraints, energy replenishment process constraints, and site capacity and congestion constraints, and to construct constraint conditions based on these constraints.