An unmanned aerial vehicle multi-target task scheduling method based on distributed cooperative game

Abstract

The application discloses a kind of unmanned aerial vehicle multi-target task scheduling methods based on distributed cooperative game, comprising the following steps: S1, define unmanned aerial vehicle state information, generate basic data;S2, construct task competition game model;S3, individual task optimization is carried out, and improved artificial native animal optimization algorithm is used;S4, improved heron colony optimization algorithm is used to carry out global task scheduling optimization;S5, adaptive adjustment in task execution process;S6, perfect task scheduling method.The application significantly improves the utilization rate of unmanned aerial vehicle resources and global task income, and can be widely applied to unmanned aerial vehicle cluster task management scenes in multiple industries such as logistics distribution, disaster relief, environmental monitoring, military reconnaissance.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of unmanned aerial vehicle multi-target task scheduling, and particularly relates to an unmanned aerial vehicle multi-target task scheduling method based on distributed cooperative game. BACKGROUND

[0002] In modern unmanned aerial vehicle cluster systems, multi-target task scheduling is a key research direction, involving the optimal allocation of multiple unmanned aerial vehicles performing multiple tasks in complex environments. With the rapid development of unmanned aerial vehicle technology, unmanned aerial vehicle clusters gradually undertake more and more tasks in the fields of military reconnaissance, environmental monitoring, emergency rescue, logistics transportation, etc. The task scheduling of unmanned aerial vehicles not only needs to consider the timeliness, priority, resource allocation and other factors of the task, but also needs to be intelligently adjusted in a dynamic environment to adapt to changes in task demand. However, the existing technology still has many challenges and deficiencies in unmanned aerial vehicle multi-target task scheduling, resulting in low scheduling efficiency, limited task completion rate, and serious waste of unmanned aerial vehicle resources.

[0003] Currently, unmanned aerial vehicle task scheduling methods are mainly divided into centralized scheduling and distributed scheduling. Centralized scheduling relies on a central controller to uniformly plan tasks and allocate them to individual unmanned aerial vehicles. Although this method can optimize in the global range, it has problems such as large communication delay, heavy computing burden, poor system robustness, etc. Once the central controller fails, the entire system will face the risk of paralysis. In addition, the centralized method is difficult to adapt to dynamic environmental changes, and when unmanned aerial vehicles have insufficient power, task failures, environmental mutations, etc., it cannot quickly adjust task allocation, resulting in a decrease in task completion rate. In contrast, the distributed scheduling method improves the flexibility and adaptability of the system by allowing unmanned aerial vehicles to make autonomous decisions on task allocation, avoiding the single-point failure problem of centralized scheduling. However, existing distributed scheduling methods usually rely on simple heuristic algorithms or rule-based strategies, lack in-depth modeling of task competition, unmanned aerial vehicle cooperation, dynamic environment adaptation, etc., and are difficult to achieve optimal task allocation.

[0004] In the optimization method of unmanned aerial vehicle task scheduling, existing researches mostly use intelligent optimization algorithms such as genetic algorithm, ant colony optimization, particle swarm optimization, etc. These algorithms have good convergence and global search ability when solving complex optimization problems, but still have certain limitations in multi-target task scheduling problems. For example, traditional optimization algorithms are prone to local optimum in high-dimensional task allocation problems, and have high computational complexity, making it difficult to meet the efficient scheduling needs of unmanned aerial vehicle clusters in real-time environments. In addition, most optimization algorithms only optimize the task selection of individual unmanned aerial vehicles, lacking a global coordination mechanism, which easily leads to the problem of some tasks not being executed while other tasks being competed by multiple unmanned aerial vehicles, resulting in unbalanced task load and decreased overall system benefit.

[0005] In recent years, game theory has been gradually introduced into the field of UAV mission scheduling, achieving optimal mission allocation by modeling the competition and cooperation relationships among UAVs. However, existing game theory-based mission scheduling methods typically employ static game models, failing to fully consider the dynamic decision-making process of UAVs and the real-time changes in the mission environment. Furthermore, most game theory methods focus only on maximizing individual payoffs, neglecting the equilibrium of global mission allocation, making it difficult to achieve optimal overall system payoffs. Simultaneously, existing game theory methods have high computational complexity, resulting in excessive computational costs in large-scale UAV swarm mission scheduling, impacting the system's real-time performance and scalability.

[0006] To address the shortcomings of existing UAV multi-objective task scheduling methods, this invention proposes a distributed cooperative game theory-based method. A task competition game model is constructed, combining an improved artificial protozoan optimization algorithm and a heron flock optimization algorithm to achieve individual UAV task optimization and global task collaborative optimization. This invention, by constructing a game theory-based distributed task competition mechanism, enables UAVs to autonomously decide on task selection based on local information. Task negotiation and payoff balancing strategies reduce task conflicts, improving the fairness and global payoff of task scheduling. To further enhance the intelligence and adaptability of task scheduling, this invention improves the artificial protozoan optimization algorithm by introducing a reinforced dormancy strategy and an adaptive evolutionary reproduction strategy. This allows UAVs to dynamically adjust their task selection strategy based on task competition and energy consumption, improving task execution success rate. Simultaneously, to address the problem of uneven global task allocation, this invention, based on the task competition game model and combined with the improved heron flock optimization algorithm, constructs a global task fitness function to measure the adaptability of the UAV group to tasks. A global task scheduling priority is constructed using a task urgency weighting factor to ensure that high-priority tasks are executed first. Furthermore, this invention proposes an adaptive task adjustment mechanism based on evolutionary game theory. During task execution, if the UAV encounters environmental changes (such as insufficient power or task failure), the task adjustment strategy is automatically triggered, and the task allocation is re-optimized based on the global payoff function until the system reaches the optimal equilibrium state.

[0007] In summary, the existing multi-target task scheduling method of unmanned aerial vehicle has the problems of poor robustness of centralized method, lack of task coordination of distributed method, high computational complexity of traditional optimization algorithm, and insufficient consideration of global optimization of game method, which is difficult to meet the demand of unmanned aerial vehicle task scheduling in complex dynamic environment. The present application builds a distributed collaborative game model, and combines an improved intelligent optimization algorithm to propose a more adaptive and efficient multi-target task scheduling method of unmanned aerial vehicle, which realizes the balanced optimization of task allocation, intelligent avoidance of task conflict, autonomous optimization of unmanned aerial vehicle task selection, and adaptive adjustment of task in dynamic environment, and significantly improves the task completion rate and system benefit of unmanned aerial vehicle cluster.

[0008] Therefore, how to provide a multi-target task scheduling method of unmanned aerial vehicle based on distributed collaborative game is a problem that those skilled in the art need to solve. SUMMARY

[0009] One object of the present application is to provide a multi-target task scheduling method of unmanned aerial vehicle based on distributed collaborative game, which has the advantages of high task scheduling balance, strong dynamic environment adaptability, high task completion rate and high unmanned aerial vehicle resource utilization rate.

[0010] According to the multi-target task scheduling method of unmanned aerial vehicle based on distributed collaborative game, the following steps are included:

[0011] S1, a task scheduling model of unmanned aerial vehicle is built, a unmanned aerial vehicle cluster and a task set are defined, a task attribute is set, a state information of unmanned aerial vehicle is initialized, and basic data is generated;

[0012] S2, based on the distributed collaborative game mechanism, the strategy set, the profit function and the game rule of the unmanned aerial vehicle agent are defined for the basic data, and a task competition game model is built, which is the core basis for the task allocation and strategy interaction of unmanned aerial vehicle;

[0013] S3, in the task competition game model, each unmanned aerial vehicle makes individual decision based on local information through the improved artificial native animal optimization algorithm, evaluates the profit of each task by using the profit function, selects the optimal task according to the strategy set, strengthens the hibernation strategy, and adaptively evolves the breeding strategy;

[0014] S4, the task competition game model further guides the task negotiation between unmanned aerial vehicles, performs global task allocation according to the game rule through the improved heron swarm optimization algorithm, measures the adaptability of unmanned aerial vehicle group to the task based on the global task fitness function, introduces a task urgency weighting factor to build a global task scheduling priority; a task allocation adjustment equation is built in the global task optimization process, a global task profit function of unmanned aerial vehicle group is set, and a task conflict factor is combined to optimize the unmanned aerial vehicle task scheduling scheme;

[0015] S5, in the task execution process, based on the game rules, strategy set and payoff function defined in the task competition game model, if the environment changes, the unmanned aerial vehicle triggers the adaptive adjustment mechanism, re-evaluates according to the payoff function and adjusts the task allocation according to the strategy set, until the system reaches the optimal equilibrium state again;

[0016] S6, after the completion of task execution, according to the long-term revenue feedback, the task competition game model is used to analyze the unmanned aerial vehicle strategy adjustment data, and the complete task scheduling method is obtained by updating the payoff function parameters.

[0017] Optionally, the S2 specifically comprises:

[0018] S21, define the unmanned aerial vehicle agent set U and the task set T, establish the distributed game framework of unmanned aerial vehicle task scheduling, set each unmanned aerial vehicle U i as the game subject, the strategy set Π is used to select the task allocation scheme, the individual unmanned aerial vehicle U i executes the task T j , the payoff function R(U i , T j ) is used to calculate the optimal strategy of each unmanned aerial vehicle, and the task competition game model G=(U, T, Π, R) is defined.

[0019] S22, set the basic rules of task competition game, each unmanned aerial vehicle U i only can obtain the task state information of itself and the adjacent unmanned aerial vehicles, and carries out local optimal strategy selection, if two or more unmanned aerial vehicles compete for the same task T j , the optimal allocation strategy is calculated according to the payoff function, and the unmanned aerial vehicle can dynamically adjust the strategy in the task execution process, and the payoff function gradually converges to the game equilibrium state.

[0020] S23, in the task competition process, set the task selection strategy, finally build the unmanned aerial vehicle task competition game model, the unmanned aerial vehicle makes autonomous decision based on the distributed cooperative game mechanism in the task allocation process, and realizes the balanced optimization of task allocation.

[0021] Optionally, the S3 specifically comprises:

[0022] S31, on the basis of the task competition game model G, each unmanned aerial vehicle U i makes individual decision based on local information, and uses the improved artificial native animal optimization algorithm for task selection, the improvement includes strengthening the hibernation strategy and the adaptive evolution and reproduction strategy.

[0023] S32, in the unmanned aerial vehicle U iWhen selecting tasks, if there is competition for task resources, insufficient power or communication obstruction, a reinforcement hibernation strategy is executed to calculate the optimal time for the UAV to enter hibernation, and the hibernation decision is optimized through the following two steps, specifically including:

[0024] S321, dynamically adjust the hibernation threshold based on energy consumption, define the UAV energy threshold function:

[0025]

[0026] Wherein, E min is the minimum energy requirement for the UAV task execution, E max is the maximum energy consumption that the UAV can withstand, is the average energy consumption of the historical task execution of the UAV, sigma(E, t) is the energy fluctuation function, which calculates the energy consumption uncertainty in the task environment, gamma1, gamma2, gamma3, gamma4 are adjustment coefficients, satisfying gamma1+gamma2+gamma3+gamma4=1;

[0027] S322, strengthen the hibernation strategy based on the task competition intensity, define the task competition factor:

[0028]

[0029] Wherein, n conf (T j ,t) represents the number of UAVs competing for task T j at time t, n avail (T j ,t) represents the number of remaining tasks that can be selected at time t;

[0030] S33, set the hibernation probability function of the UAV:

[0031] When E(U i ,t) < E th (U i ,t), P sleep (U i ,t) = 1;

[0032] Wherein, P sleep (U i ,t) is the probability of the UAV entering hibernation, tau is the temperature parameter, when the energy E(U i ,t) of the UAV is lower than the threshold E th (U i ,t), the hibernation probability is set to 1;

[0033] S34, in the process of selecting the task of the UAV, an adaptive evolution breeding strategy is introduced, the historical task execution data of the UAV is used to optimize the task allocation strategy, and an individual evolution equation is set;

[0034]

[0035] wherein, is the strategy of the UAV after the selection of the task in the t+1 round, δ is an evolution adjustment factor, and the strategy updating step is controlled, is a gradient of the task benefit function to the strategy, ρ is an experience learning factor, and the historical task execution experience is used to optimize the strategy, θ is a task environment adaptation factor, and the strategy is dynamically adjusted in combination with time, is an intelligent evolution function, the strategy variation direction is adjusted according to the task execution success rate and the task difficulty, so that the UAV evolves a more stable task selection strategy, and ζ·(E th (U i ,t)-E(U i ,t)) is an influence item for the UAV to evolve a scheduling strategy more inclined to low-energy-consumption tasks in a low-energy state;

[0036] S35, a task selection mechanism based on dynamic energy constraint is introduced, the UAV with low energy will automatically reduce the benefit weight of the high-energy-consumption task, and prefer to select the low-energy-consumption task:

[0037] R adj (U i ,T j )=R(U i ,T j )-λ E ·(E(U i ,t)-E th (U i ,t));

[0038] wherein, R(U i ,T j ) is a benefit function of the individual UAV U i executing the task T j , R adj (U i ,T j ) is an adjusted benefit function of the individual UAV U i executing the task T j , and λ E is an energy adjustment factor;

[0039] S36, in the process of executing the task, the unmanned aerial vehicle combines the reinforced dormancy strategy and the adaptive evolution and reproduction strategy, dynamically adjusts the task allocation scheme of itself, if the unmanned aerial vehicle is in the dormant state for a long time, the task selection strategy is adjusted again according to the task benefit function, in the task competition process, if two or more unmanned aerial vehicles compete for the same task T j , the optimal task allocation is calculated based on the game rules in the task competition game model, and the task priority is dynamically adjusted.

[0040] Optionally, the S4 specifically comprises:

[0041] S41, on the basis of the individual task selection of the improved artificial native animal optimization algorithm, the improved heron colony optimization algorithm is used for global task allocation optimization;

[0042] S42, in the process of task competition game, define the global task fitness evaluation function, measure the adaptability of the whole unmanned aerial vehicle cluster to the task, set the global fitness function:

[0043]

[0044] Wherein, F global (T,t) is the total fitness of the unmanned aerial vehicle group in executing the task set T at time t, F task (U i ,T j ,t) is the fitness of the individual unmanned aerial vehicle to the task T j ;

[0045] S43, based on the global fitness function optimization task allocation, introduce the task urgency weighting factor P urg (T j ,t), construct the global task scheduling priority:

[0046]

[0047] Wherein, θ1 is the maximum value of the task urgency, θ2 is the emergency rate factor, t d is the deadline of the task T j , t is the task execution time, P alloc (T j ,t) represents the priority weight of the task T j in the global task allocation;

[0048] S44, in the process of global task optimization, a dynamic task cooperation mechanism is used to construct the global task allocation scheme:

[0049]

[0050] Wherein, Ω t+1Ω represents a global task allocation scheme of the UAV group at time t+1 t Ω represents a global task allocation scheme of the UAV group at time t, a is a global fitness optimization step, and β is a task allocation priority optimization step, represents a gradient function;

[0051] S45, set a global task benefit function of the UAV group:

[0052]

[0053] wherein, R global (t) is the total benefit of the UAV task allocation at time t, C conf (U i ,T j ,t) is a task conflict factor between UAVs, λ C is a task conflict adjustment factor.

[0054] Optionally, the S5 specifically comprises:

[0055] S51, set a UAV self-adaptive adjustment mechanism to trigger a task adjustment strategy according to real-time environmental information in the task execution process;

[0056] S52, define a UAV task execution state function to monitor the state of the UAV in the task execution process:

[0057] S exec (U i ,T j ,t)=α1·E rem (U i ,t)+α2·R(U i ,T j )+α3·C conf (U i ,T j ,t);

[0058] wherein, S exec (U i ,T j ,t) is the state value of the UAV executing the task T j at time t, E rem (U i ,t) is the remaining power of the UAV, R(U i ,T j ) is the task benefit of the individual UAV U i executing the task T j , and C conf (U i ,T jt) is an inter-unmanned aerial vehicle task conflict factor, alpha1, alpha2, alpha3 are task state weight coefficients, and alpha1+alpha2+alpha3=1 is satisfied;

[0059] S53, a task state threshold is set, and an adaptive adjustment mechanism is triggered according to the unmanned aerial vehicle task state;

[0060]

[0061] Wherein, delta L is a low threshold of the task state, if the unmanned aerial vehicle task state evaluation is lower than the value, the adjustment strategy is triggered;

[0062] S54, after the unmanned aerial vehicle triggers the adaptive adjustment mechanism, global task reassignment is carried out based on the improved heron swarm optimization algorithm, the unmanned aerial vehicle task execution strategy is adjusted, and the individual task selection scheme is optimized according to the improved artificial native animal optimization algorithm, until the system reaches the optimal equilibrium state again, and finally the optimized task competition game model is defined as:

[0063] G * =(U,T,Π * ,R * );

[0064] Wherein, U is the unmanned aerial vehicle intelligent agent set, T is the task set, Pi * is the final optimized unmanned aerial vehicle task allocation strategy set, R * is the final optimized unmanned aerial vehicle task income function.

[0065] The beneficial effects of the present application are:

[0066] (1) the present application constructs a task competition game model based on a distributed cooperative game mechanism, defines the income function, strategy set and game rules among unmanned aerial vehicle intelligent agents, so that the unmanned aerial vehicle can make autonomous decision on task selection on the basis of local information, and reduce task conflicts through task negotiation and income equilibrium strategy. In addition, combined with the improved heron swarm optimization algorithm (IESOA), the present application introduces a global task fitness function to measure the adaptability of the unmanned aerial vehicle group to the task, and combines the task urgency weighting factor to construct the global task scheduling priority, so as to ensure that high priority tasks are executed preferentially, thereby improving the balance of task allocation and avoiding the problem of unmanned execution or waste of unmanned aerial vehicle resources.

[0067] (2) The application introduces an adaptive task adjustment mechanism based on evolutionary game in the task execution process. The improved artificial native animal optimization algorithm (APO) is used to optimize the individual task of the unmanned aerial vehicle. The hibernation strategy and adaptive evolution breeding strategy are set to enable the unmanned aerial vehicle to dynamically adjust the task selection strategy according to the task competition situation, energy state and task execution difficulty, thereby improving the task execution success rate. In addition, if the unmanned aerial vehicle encounters power shortage, task failure, task demand change and other environmental changes during task execution, the adaptive adjustment mechanism of the application can trigger the task reassignment strategy, and perform global task optimization according to the benefit function and strategy set, so as to ensure that the unmanned aerial vehicle task scheduling scheme can still maintain the optimal balance in a complex environment, and improve the stability and task completion rate of the system.

[0068] (3) The application combines the improved artificial native animal optimization algorithm (APO) and the improved heron eagle search optimization algorithm (IESOA) to propose a scheduling method combining individual decision optimization and global task scheduling optimization, so that the unmanned aerial vehicle group can efficiently complete the task. The APO algorithm introduces a reinforced hibernation strategy and an adaptive evolution breeding strategy to ensure that the unmanned aerial vehicle can autonomously optimize itself according to its power, task benefit and task competition situation, thereby improving the task adaptation degree. The IESOA algorithm performs task coordination optimization, global task adaptation degree evaluation, task urgency control and benefit function optimization to ensure the global optimality of the task scheduling and improve the system task execution efficiency. In addition, the benefit balance mechanism of the application effectively reduces the imbalance phenomenon in task execution, so that the unmanned aerial vehicle resources are fully utilized, and the task completion efficiency and benefit level of the entire unmanned aerial vehicle cluster are improved. BRIEF DESCRIPTION OF DRAWINGS

[0069] The accompanying drawings are included to provide a further understanding of the application, and constitute a part of the specification, illustrate embodiments of the application, and are used to explain the application, and do not constitute a limitation of the application. In the drawings:

[0070] Figure 1 The application proposes a whole flow chart of a multi-target task scheduling method for unmanned aerial vehicles based on distributed cooperative game. DETAILED DESCRIPTION

[0071] The application will now be described in further detail with reference to the accompanying drawings. These drawings are simplified schematic diagrams, and only schematically show the basic structure of the application, and therefore only show the components related to the application.

[0072] REFERENCE Figure 1 A multi-target task scheduling method for unmanned aerial vehicles based on distributed cooperative game includes the following steps:

[0073] S1, a UAV task scheduling model is constructed, a UAV cluster and a task set are defined, task attributes are set, state information of the UAV is initialized, and basic data is generated;

[0074] S2, based on a distributed cooperative game mechanism, a strategy set, a payoff function and a game rule of a UAV agent are defined for the basic data, a task competition game model is constructed, and the task competition game model is used as a core basis for UAV task allocation and strategy interaction;

[0075] S3, in the task competition game model, each UAV makes individual decisions based on local information through an improved artificial primitive animal optimization algorithm, evaluates the benefits of each task using the benefit function, selects the optimal task according to the strategy set, strengthens the hibernation strategy, and adaptively evolves the breeding strategy;

[0076] S4, the task competition game model further guides the task negotiation between UAVs, performs global task allocation according to the game rules through an improved white egret optimization algorithm, measures the adaptability of the UAV group to the task based on a global task fitness function, introduces a task urgency weighting factor to construct a global task scheduling priority; in the global task optimization process, a task allocation adjustment equation is constructed, a global task benefit function of the UAV group is set, and a UAV task scheduling scheme is optimized in combination with a task conflict factor;

[0077] S5, during task execution, based on the game rules, strategy set and benefit function defined in the task competition game model, if the environment changes, the UAV triggers an adaptive adjustment mechanism, re-evaluates the task allocation according to the benefit function and adjusts the task allocation according to the strategy set until the system reaches an optimal equilibrium state again;

[0078] S6, after task execution is completed, the UAV strategy adjustment data is analyzed using the task competition game model based on long-term benefit feedback, the benefit function parameters are updated, and a complete task scheduling method is obtained.

[0079] In the embodiment, the S2 specifically includes:

[0080] S21, a UAV agent set U and a task set T are defined, a distributed game framework for UAV task scheduling is established, and each UAV U i as a game subject, the strategy set Π is used to select a task allocation scheme, and each UAV U i executes a task T j The benefit function R(U i , T j ) is used to calculate the optimal strategy of each UAV, and a task competition game model G=(U, T, Π, R) is defined;

[0081] S22, the basic rules of the task competition game are set, each UAV U iOnly the task state information of itself and neighboring UAVs can be acquired, and a local optimal strategy is selected, if two or more UAVs compete for the same task T j A competition occurs, an optimal allocation strategy is calculated according to a benefit function, and the UAV can dynamically adjust the strategy during task execution, and the benefit function gradually converges to a game equilibrium state.

[0082] S23, in the task competition process, a task selection strategy is set, and finally a UAV task competition game model is constructed, the UAV makes autonomous decision based on a distributed cooperative game mechanism during task allocation, and balanced optimization of task allocation is realized.

[0083] In this embodiment, the S3 specifically includes:

[0084] S31, on the basis of the task competition game model G, each UAV U i Based on local information, an improved artificial primitive animal optimization algorithm is used for task selection, and the improvement includes strengthening the hibernation strategy and the adaptive evolution and reproduction strategy.

[0085] S32, in the UAV U i When selecting a task, if the task resource competition, insufficient power or communication obstruction problem is faced, the strengthened hibernation strategy is executed, the optimal opportunity for the UAV to enter hibernation is calculated, and the hibernation decision is optimized through the following two steps, specifically including:

[0086] S321, the hibernation threshold is dynamically adjusted based on energy consumption, and the UAV energy threshold function is defined:

[0087]

[0088] Wherein, E min is the minimum energy requirement for UAV task execution, E max is the maximum energy consumption that the UAV can bear, is the average energy consumption of the historical task execution of the UAV, and σ(E, t) is the energy fluctuation function, which calculates the energy consumption uncertainty in the task environment, and γ1, γ2, γ3, γ4 are adjustment coefficients, satisfying γ1+γ2+γ3+γ4=1;

[0089] S322, the hibernation strategy is strengthened based on the task competition intensity, and the task competition factor is defined:

[0090]

[0091] Wherein, n conf (T j , t) represents the number of UAVs competing for the task T j at time t, and n avail (T jt) represents the optional remaining task number at time t;

[0092] S33, set the probability function of the drone hibernation:

[0093] When E(U i ,t)<E th (U i ,t), P sleep (U i ,t) = 1;

[0094] Where P sleep (U i ,t) is the probability of the drone entering hibernation, τ is the temperature parameter, when the energy E(U i ,t) of the drone is lower than the threshold E th (U i ,t), the hibernation probability is set to 1;

[0095] S34, in the process of optimizing the task selection of the drone, an adaptive evolution and breeding strategy is introduced, the historical task execution data of the drone is used to optimize the task allocation strategy, and an individual evolution equation is set:

[0096]

[0097] Where, is the strategy of the drone after t+1 rounds of task selection, δ is the evolution adjustment factor, controlling the strategy update step, is the gradient of the task benefit function to the strategy, ρ is the experience learning factor, based on the historical task execution experience to optimize the strategy, θ is the task environment adaptation factor, which adjusts the strategy dynamically combined with time, is the intelligent evolution function, which adjusts the strategy variation direction according to the task execution success rate and task difficulty, so that the drone evolves a more stable task selection strategy, ζ·(E th (U i ,t)-E(U i ,t)) is an influence term for the drone to evolve a scheduling strategy more biased towards low energy consumption tasks in a low energy state;

[0098] S35, introduce a task selection mechanism based on dynamic energy constraint, the drones with lower energy will automatically reduce the benefit weight of high energy consumption tasks, and prefer to select low energy consumption tasks:

[0099] R adj (U i ,T j ) = R(U i ,T j )-λ E ·(E(Ui t) - E th (U i ,t) ;

[0100] wherein R(U i ,T j ) is a reward function of the individual unmanned aerial vehicle U i performing the task T j , R adj (U i ,T j ) is an adjusted reward function of the individual unmanned aerial vehicle U i performing the task T j , and λ E is an energy adjustment factor;

[0101] S36, in the process of performing the task, the unmanned aerial vehicle dynamically adjusts the task allocation scheme of itself in combination with the reinforced dormancy strategy and the self-adaptive evolutionary breeding strategy, if the unmanned aerial vehicle is in a dormant state for a long time, the task selection strategy is readjusted according to the task reward function, in the process of task competition, if two or more unmanned aerial vehicles compete for the same task T j , the optimal task allocation is calculated based on the game rule in the task competition game model, and the task priority is dynamically adjusted.

[0102] In the embodiment, the S4 specifically comprises:

[0103] S41, on the basis of the improved artificial native animal optimization algorithm individual task selection, the improved heron colony optimization algorithm is used for global task allocation optimization;

[0104] S42, in the process of task competition game, a global task fitness evaluation function is defined to measure the adaptability of the whole unmanned aerial vehicle cluster to the task, and the global fitness function is set as:

[0105]

[0106] wherein F global (T,t) is the total fitness of the unmanned aerial vehicle group in performing the task set T at time t, and F task (U i ,T j ,t) is the fitness of the individual unmanned aerial vehicle to the task T j ;

[0107] S43, based on the global fitness function to optimize the task allocation, a task urgency weighting factor P urg (T j ,t) is introduced to construct the global task scheduling priority:

[0108]

[0109] wherein θ1 is the maximum value of the task urgency, θ2 is the urgency change rate factor, t d is the deadline of task T j , t is the task execution time, P alloc (T j , t) represents the task T j priority weight in the global task allocation;

[0110] S44, in the global task optimization process, a dynamic task coordination mechanism is adopted to construct a global task allocation scheme:

[0111]

[0112] wherein Ω t+1 represents the global task allocation scheme of the UAV group at time t+1, Ω t represents the global task allocation scheme of the UAV group at time t, α is the global fitness optimization step, β is the task allocation priority optimization step, represents the gradient function;

[0113] S45, set the global task benefit function of the UAV group:

[0114]

[0115] wherein R global (t) is the total benefit of the UAV task allocation at time t, C conf (U i , T j , t) is the task conflict factor between UAVs, λ C is the task conflict adjustment factor.

[0116] In the embodiment, the S5 specifically comprises:

[0117] S51, set the UAV self-adaptive adjustment mechanism, and trigger the task adjustment strategy according to the real-time environment information in the task execution process;

[0118] S52, define the UAV task execution state function, and monitor the state of the UAV in the task execution process:

[0119] S exec (U i , T j , t) = α1·E rem (U i , t) + α2·R(U i , T j ) + α3·C conf (U i , T j , t);

[0120] Among them, S exec (U i ,T j (t) represents the time t when the UAV performs the task T. j The state value, E rem (U i R(U) represents the remaining battery power of the drone. i ,T j For individual drones U i Execute task T j Task rewards, C conf (U i ,T j ,t) is the task conflict factor between UAVs, and α1,α2,α3 are the task state weight coefficients, satisfying α1+α2+α3=1;

[0121] S53. Set task status thresholds and trigger an adaptive adjustment mechanism based on the UAV task status:

[0122]

[0123] Where, δ L The task status is set as a low threshold. If the drone's task status assessment is lower than this value, an adjustment strategy will be triggered.

[0124] S54. After the UAV triggers the adaptive adjustment mechanism, a global task reallocation is performed based on the improved egret flock optimization algorithm, the UAV task execution strategy is adjusted, and the individual task selection scheme is optimized based on the improved artificial protozoan optimization algorithm until the system reaches the optimal equilibrium state again. The final optimized task competition game model is defined as follows:

[0125] G * =(U,T,Π) * ,R * );

[0126] Where U represents the set of unmanned aerial vehicle (UAV) intelligent agents, T represents the set of tasks, and Π * R assigns a set of strategies to the final optimized set of drone missions. * This is the final optimized drone mission reward function.

[0127] Example 1:

[0128] To verify the feasibility of the present application in implementation, the present application is applied to a certain unmanned aerial vehicle cluster logistics distribution system, which is used for urban rapid distribution scenarios, involves multiple unmanned aerial vehicles performing multiple distribution tasks, optimizes the unmanned aerial vehicle scheduling scheme in a dynamic environment, and improves the task completion rate and system revenue. The main goal of the experiment is to verify the advantages of the present application in task scheduling balance, dynamic environment adaptability, task completion rate, and unmanned aerial vehicle resource utilization. The experimental site is set in a certain large logistics center, which needs to schedule hundreds of unmanned aerial vehicles every day to perform cross-city distribution tasks, involves multi-objective task allocation, different task priorities, and time constraints for some tasks.

[0129] The experiment adopts two schemes for comparison, one is the traditional unmanned aerial vehicle task scheduling algorithm based on heuristic rules, and the other is the unmanned aerial vehicle multi-objective task scheduling method based on distributed cooperative game proposed by the present application. In the experiment, 50 unmanned aerial vehicles are set, and the number of distribution tasks is 200, 400, 600, 800, and 1000 respectively, to test the scheduling effect of the system under different task loads. At the same time, considering the influence of dynamic changes of tasks, 10%-30% of the tasks are randomly set to be adjusted due to environmental changes (such as weather changes, traffic control, user change requirements) in the experiment, to test the adaptability of the scheduling algorithm to dynamic environment.

[0130] In the experiment, the scheduling method of the present application first constructs a task competition game mechanism based on the distributed cooperative game model, so that the unmanned aerial vehicle can autonomously select the optimal task based on local information. Then, combined with the improved artificial native animal optimization algorithm, the unmanned aerial vehicle individual task is optimized, so that the unmanned aerial vehicle can dynamically adjust the task selection scheme according to its remaining power, task revenue and task competition, and optimize the task execution through the hibernation strategy and adaptive evolution breeding strategy. At the same time, the improved heron swarm optimization algorithm is used for global task scheduling, so that the task allocation is more balanced, the task conflict between unmanned aerial vehicles is reduced, and the overall system revenue is improved. When the unmanned aerial vehicle encounters an emergency event (such as weather change, user order modification, unmanned aerial vehicle failure, etc.) during task execution, the adaptive adjustment mechanism of the present application can trigger task reallocation, dynamically optimize according to the task revenue function and strategy set, so that the system quickly recovers to the optimal balanced state.

[0131] The experimental results show that the unmanned aerial vehicle task scheduling method has obvious advantages over the traditional scheduling method in terms of task completion rate, unmanned aerial vehicle task load balancing degree, unmanned aerial vehicle resource utilization rate, dynamic environment adaptability and system benefit. When the total number of tasks is 1000, the task completion rate of the present application reaches 95.2%, which is 12.8% higher than that of the traditional method; the task load balancing degree of the unmanned aerial vehicle is improved by 20.3%, that is, the task allocation among different unmanned aerial vehicles is more balanced, avoiding the situation that some unmanned aerial vehicles are overloaded or idle; the resource utilization rate of the unmanned aerial vehicle is improved by 18.7%, that is, the power consumption of the unmanned aerial vehicle is more reasonable, so that the overall system consumes less energy while performing more tasks. In addition, in the case that 30% of the tasks change dynamically, the scheduling method of the present application can complete task reallocation within an average of 3.7 seconds, while the traditional method needs 7.9 seconds, proving that the scheduling method of the present application has better adaptability in dynamic environment.

[0132] Table 1: Unmanned aerial vehicle task scheduling experimental data table

[0133]

[0134] From the experimental data, it can be seen that the task completion rate of the present application is always higher than that of the traditional method, and the difference is more significant when the number of tasks is large (600 or more). The task completion rate of the present application remains at 95.2% when the number of tasks is 1000, while the traditional method decreases to 82.4%, which shows that the scheduling method of the present application has more advantages in high-load task scenarios.

[0135] In addition, the task load balancing degree reflects the uniformity of task allocation of the unmanned aerial vehicle, and the load balancing degree of the present application reaches 85.6% when the number of tasks is 1000, which is 16.4% higher than that of the traditional method, indicating that the optimization strategy of the present application can effectively reduce the imbalance problem of task allocation and avoid the situation that some unmanned aerial vehicles work overload while other unmanned aerial vehicles are idle.

[0136] In terms of resource utilization rate, the unmanned aerial vehicle task scheduling method of the present application optimizes the task execution path of the unmanned aerial vehicle, so that the power consumption of the unmanned aerial vehicle is more balanced, and the overall resource utilization rate is improved by 18.7%, which shows that the present application can effectively reduce the waste of unmanned aerial vehicle resources and improve the task execution efficiency. In addition, when the number of tasks reaches 1000, the average scheduling time of the present application is only 3.7 seconds, which is 4.2 seconds less than that of the traditional method, which shows that the present application has higher scheduling response capability in dynamic environment and can quickly complete task allocation adjustment.

[0137] The unmanned aerial vehicle multi-target task scheduling method based on distributed cooperative game of the present application is superior to the traditional method in key indicators such as task completion rate, load balancing degree, resource utilization rate and scheduling time, and proves the significant advantages of the present application in improving task execution efficiency, reducing task conflicts, improving unmanned aerial vehicle resource utilization rate and the like, and can be more effectively applied to large-scale unmanned aerial vehicle cluster task scheduling in a dynamic environment.

[0138] The above merely describes a preferred specific embodiment of the present application, but the protection scope of the present application is not limited thereto, any person skilled in the art can make equivalent replacements or changes within the technical range disclosed by the present application according to the technical solution and inventive concept of the present application, which should be covered within the protection scope of the present application.

Claims

1. A method for multi-target task scheduling of unmanned aerial vehicles based on distributed cooperative game, characterized in that, The method comprises the following steps: S1, defining a UAV cluster and a task set, initializing state information of the UAVs, and generating basic data; S2, defining a strategy set, a payoff function, and a game rule of the UAV agents based on a distributed cooperative game mechanism, and constructing a task competition game model based on the basic data; S3, in the task competition game model, each UAV makes individual decisions based on local information through an improved artificial naive animal optimization algorithm, evaluates the payoff of each task using the payoff function, selects the optimal task according to the strategy set, strengthens the hibernation strategy, and adaptively evolves the reproduction strategy; S4, the task competition game model performs global task allocation through an improved heron swarm optimization algorithm, measures the adaptability of the UAV cluster to the tasks based on a global task fitness function, introduces a task urgency weighting factor to construct a global task scheduling priority, constructs a task allocation adjustment equation in the global task optimization process, sets a global task payoff function of the UAV cluster, and optimizes the UAV task scheduling scheme in combination with a task conflict factor; S5, an adaptive adjustment mechanism is constructed, the task allocation is adjusted according to the payoff function evaluation and the strategy set, and the system is re-optimized until the optimal equilibrium state is reached; S6, after the task execution is completed, the payoff function parameters are updated according to the long-term payoff feedback, and a complete task scheduling method is obtained.

2. The method of claim 1, wherein, The S2 specifically comprises: S21, define the set of unmanned aerial vehicle agents U and the set of tasks T, establish a distributed game framework for unmanned aerial vehicle task scheduling, set each unmanned aerial vehicle U i As a game subject, the strategy set Π is used to select the task allocation scheme, and the individual unmanned aerial vehicle U i Executes task T j The revenue function R(U i ,T j ) is used to calculate the optimal strategy of each unmanned aerial vehicle, and a task competition game model G=(U, T, Π, R) is defined. S22, set the basic rules of the task competition game, each unmanned aerial vehicle U i Only the task state information of itself and adjacent unmanned aerial vehicles can be obtained, and a local optimal strategy is selected, if two or more unmanned aerial vehicles compete for the same task T j The optimal allocation strategy is calculated according to the benefit function, and the unmanned aerial vehicle can dynamically adjust the strategy during the task execution process, and the benefit function gradually converges to the game equilibrium state. S23, in the task competition process, a task selection strategy is set, and finally a UAV task competition game model is constructed; the UAV makes autonomous decisions based on the distributed cooperative game mechanism in the task allocation process, and the balanced optimization of task allocation is realized.

3. The method of claim 1, wherein, The S3 specifically comprises: S31、On the basis of task competition game model G, each UAV U i Based on local information, individual decision-making, improved artificial native animal optimization algorithm is used for task selection, including strengthening the hibernation strategy and adaptive evolution breeding strategy. S32、In the unmanned aerial vehicle U i When selecting a task, if the task resource competition, insufficient power or communication blockage problem is encountered, the reinforcement sleep strategy is executed, the optimal opportunity for the unmanned aerial vehicle to enter sleep is calculated, and the sleep decision is optimized through the following two steps, specifically including: S321, the hibernation threshold is dynamically adjusted based on energy consumption, and a UAV energy threshold function is defined: wherein E min is the minimum energy requirement for the UAV to perform the task, max is the maximum energy consumption that the UAV can withstand, is the average energy consumption of the UAV in historical task execution, σ(E,t) is an energy fluctuation function that calculates the uncertainty of energy consumption in the task environment, γ1, γ2, γ3, γ4 are adjustment coefficients that satisfy γ1+γ2+γ3+γ4=1; S322, the hibernation strategy is strengthened based on the task competition intensity, and a task competition factor is defined: wherein n conf (T j (T j the number of competing drones, n avail (T j the number of optional remaining tasks at time t; S33, a hibernation probability function of the UAV is set: When E(U i ,t) < E th (U i ,t), P sleep (U i ,t) = 1; Among them, P sleep (U i ,t) represents the probability of the drone entering hibernation, τ is the temperature parameter, and when the drone's energy E(U i ,t) is below the threshold E th (U i When t), the sleep probability is set to 1; S34, in the UAV task selection optimization process, an adaptive evolution and reproduction strategy is introduced, the task allocation strategy is optimized according to the historical task execution data of the UAV, and an individual evolution equation is set: wherein, is the policy of the UAV after selecting the task in the t+1 round, δ is the evolution adjustment factor, and controls the policy update step, is the gradient of the task reward function to the policy, and ρ is the experience learning factor, which is based on the historical task execution experience is the policy optimization, and θ is the task environment adaptation factor, which dynamically adjusts the policy in combination with time, is the intelligent evolution function, which adjusts the policy mutation direction according to the task execution success rate and the task difficulty, so that the UAV evolves a more stable task selection policy, and ζ·(E th (U i ,t)-E(U i ,t)) is an influence term for the UAV to evolve a more biased low-energy task scheduling policy in a low-energy state; S35, a task selection mechanism based on dynamic energy constraint is introduced, the UAV with low energy will automatically reduce the payoff weight of high energy consumption tasks, and prefer to select low energy consumption tasks: R adj (U i ,T j )=R(U i ,T j )-λ E ·(E(U i ,t)-E th (U i ,t)); wherein R(U i ,T j ) is a reward function of the individual unmanned aerial vehicle U i performing the task T j , R adj (U i ,T j ) is an adjusted reward function of the individual unmanned aerial vehicle U i performing the task T j , and λ E is an energy adjustment factor; S36, in the process of executing the task, the unmanned aerial vehicle dynamically adjusts the task allocation scheme of itself in combination with the reinforced dormancy strategy and the adaptive evolution and breeding strategy, if the unmanned aerial vehicle is in the dormant state for a long time, the task selection strategy is adjusted again according to the task benefit function, in the process of task competition, if two or more unmanned aerial vehicles compete for the same task T j , the optimal task allocation is calculated based on the game rules in the task competition game model, and the task priority is dynamically adjusted.

4. The method of claim 1, wherein, The S4 specifically comprises: S41, on the basis of individual task selection of the improved artificial naive animal optimization algorithm, the improved heron swarm optimization algorithm is used for global task allocation optimization; S42, in the task competition game process, a global task fitness evaluation function is defined to measure the adaptability of the entire UAV cluster to the tasks, and a global fitness function is set: wherein F global (T,t) is the total fitness of the UAV group in performing the task set T at time t, F task (U i ,T j ,t) is the fitness of the individual UAV to the task T j at time t. S43, optimizing task allocation based on global fitness function, introducing task urgency weighting factor P urg (T j , t), constructing global task scheduling priority: where θ1 is the maximum value of the task urgency, θ2 is the urgency change rate factor, t d is the deadline of task T j , t is the task execution time, P alloc (T j , t) represents the priority weight of task T j in the global task allocation; S44, in the global task optimization process, a dynamic task cooperation mechanism is used to construct a global task allocation scheme: wherein, Ω t+1 represents the global task allocation scheme of the UAV group at time t+1, Ω t represents the global task allocation scheme of the UAV group at time t, a is the global fitness optimization step, β is the task allocation priority optimization step, represents the gradient function; S45, a global task payoff function of the UAV cluster is set: wherein R global (t) is the total benefit of the UAV task assignment at time t, C conf (U i , T j , t) is the task conflict factor between UAVs, λ C is the task conflict adjustment factor.

5. The method of claim 1, wherein, The S5 specifically comprises: S51, a UAV adaptive adjustment mechanism is set, and the task adjustment strategy is triggered according to real-time environmental information in the task execution process; S52, a UAV task execution state function is defined to monitor the state of the UAV in the task execution process: S exec (U i ,T j ,t) = α1·E rem (U i ,t) + α2·R(U i ,T j ) + α3·C conf (U i ,T j ,t). Among them, S exec (U i ,T j (t) represents the time t when the UAV performs the task T. j The state value, E rem (U i R(U) represents the remaining battery power of the drone. i ,T j For individual drones U i Execute task T j Task rewards, C conf (U i ,T j ,t) is the task conflict factor between UAVs, and α1,α2,α3 are the task state weight coefficients, satisfying α1+α2+α3=1; S53, a task state threshold is set, and the adaptive adjustment mechanism is triggered according to the UAV task state: wherein δ L is a low threshold value for the mission state, below which the adjustment strategy is triggered if the mission state of the UAV is assessed. S54, after the UAV triggers the adaptive adjustment mechanism, the global task is redistributed based on the improved white egret colony optimization algorithm, the UAV task execution strategy is adjusted, and the individual task selection scheme is optimized according to the improved artificial native animal optimization algorithm, until the system reaches the optimal equilibrium state again, and the final optimized task competition game model is defined as: G * = (U, T, Π * , R * ); Wherein, U is a set of unmanned aerial vehicle agents, T is a set of tasks, Π * is a set of final optimized unmanned aerial vehicle task allocation strategies, R * is a final optimized unmanned aerial vehicle task revenue function.

Images