A cooperative control method of multi-class UAV(unmanned air vehicle) team is investigated.During the mission,the UAVs perform search,classification,attack and battle damage assessment(BDA) tasks at various locations,...A cooperative control method of multi-class UAV(unmanned air vehicle) team is investigated.During the mission,the UAVs perform search,classification,attack and battle damage assessment(BDA) tasks at various locations,which involves a combination of the team intelligence type of decision making combined with control,estimate and real-time trajectory optimization.The search-theoretic approach based on rate of return(ROR) maps is developed to get the cooperative search strategy.Templates are developed and views are combined to maximize the probability of correct target identification over various aspect angles.Monte Carle simulation runs for the scenario to evaluate the performance of the approach with various decision parameters,UAVs distributions and UAV team characteristics.Simulation results show that the cooperative behavior can significantly improve the operational effectiveness of UAV team,and the cooperative control allows for near optimal solution of the correlative behavior of a group of UAVs in battlefield.展开更多
The overall performance of multi-robot collaborative systems is significantly affected by the multi-robot task allocation.To improve the effectiveness,robustness,and safety of multi-robot collaborative systems,a multi...The overall performance of multi-robot collaborative systems is significantly affected by the multi-robot task allocation.To improve the effectiveness,robustness,and safety of multi-robot collaborative systems,a multimodal multi-objective evolutionary algorithm based on deep reinforcement learning is proposed in this paper.The improved multimodal multi-objective evolutionary algorithm is used to solve multi-robot task allo-cation problems.Moreover,a deep reinforcement learning strategy is used in the last generation to provide a high-quality path for each assigned robot via an end-to-end manner.Comparisons with three popular multimodal multi-objective evolutionary algorithms on three different scenarios of multi-robot task allocation problems are carried out to verify the performance of the proposed algorithm.The experimental test results show that the proposed algorithm can generate sufficient equivalent schemes to improve the availability and robustness of multi-robot collaborative systems in uncertain environments,and also produce the best scheme to improve the overall task execution efficiency of multi-robot collaborative systems.展开更多
A new method for a cooperative multi-task allocation problem(CMTAP) is proposed in this paper,taking into account the multi-ship, multi-target, multi-task and multi-constraint characteristics in a multi-ship cooperati...A new method for a cooperative multi-task allocation problem(CMTAP) is proposed in this paper,taking into account the multi-ship, multi-target, multi-task and multi-constraint characteristics in a multi-ship cooperative driving(MCD) system. On the basis of the general CMTAP model, an MCD task assignment model is established. Furthermore, a genetic ant colony hybrid algorithm(GACHA) is proposed for this model using constraints, including timing constraints, multi-ship collaboration constraints and ship capacity constraints. This algorithm uses a genetic algorithm(GA) based on a task sequence, while the crossover and mutation operators are based on similar tasks. In order to reduce the dependence of the GA on the initial population, an ant colony algorithm(ACA) is used to produce the initial population. In order to meet the environmental constraints of ship navigation, the results of the task allocation and path planning are combined to generate an MCD task planning scheme. The results of a simulated experiment using simulated data show that the proposed method can make the assignment more optimized on the basis of satisfying the task assignment constraints and the ship navigation environment constraints. Moreover, the experimental results using real data also indicate that the proposed method can find the optimal solution rapidly, and thus improve the task allocation efficiency.展开更多
在网络状态和任务需求的动态变化下,为减少模型推理时延和计算成本,在软件定义网络(Software Defined Network,SDN)中提出了一种基于模型划分的云边协同推理算法。通过构建复杂度预测器分配任务执行环境,采用深度Q网络(Deep Q-network,D...在网络状态和任务需求的动态变化下,为减少模型推理时延和计算成本,在软件定义网络(Software Defined Network,SDN)中提出了一种基于模型划分的云边协同推理算法。通过构建复杂度预测器分配任务执行环境,采用深度Q网络(Deep Q-network,DQN)算法对边缘环境中的推理模型进行自适应划分与卸载;以及用SDN技术从全局感知推理任务与网络资源,实现动态网络环境下网络资源的合理分配。试验结果表明:SDN中基于模型划分的云边协同推理算法具有良好的收敛能力,在动态环境中具有较好的鲁棒性。与现有的推理算法相比,该算法能够在合理分配计算资源的前提下,满足协同推理低时延的目标要求。展开更多
文摘A cooperative control method of multi-class UAV(unmanned air vehicle) team is investigated.During the mission,the UAVs perform search,classification,attack and battle damage assessment(BDA) tasks at various locations,which involves a combination of the team intelligence type of decision making combined with control,estimate and real-time trajectory optimization.The search-theoretic approach based on rate of return(ROR) maps is developed to get the cooperative search strategy.Templates are developed and views are combined to maximize the probability of correct target identification over various aspect angles.Monte Carle simulation runs for the scenario to evaluate the performance of the approach with various decision parameters,UAVs distributions and UAV team characteristics.Simulation results show that the cooperative behavior can significantly improve the operational effectiveness of UAV team,and the cooperative control allows for near optimal solution of the correlative behavior of a group of UAVs in battlefield.
基金the Shanghai Pujiang Program (No.22PJD030),the National Natural Science Foundation of China (Nos.61603244 and 71904116)the National Natural Science Foundation of China-Shandong Joint Fund (No.U2006228)。
文摘The overall performance of multi-robot collaborative systems is significantly affected by the multi-robot task allocation.To improve the effectiveness,robustness,and safety of multi-robot collaborative systems,a multimodal multi-objective evolutionary algorithm based on deep reinforcement learning is proposed in this paper.The improved multimodal multi-objective evolutionary algorithm is used to solve multi-robot task allo-cation problems.Moreover,a deep reinforcement learning strategy is used in the last generation to provide a high-quality path for each assigned robot via an end-to-end manner.Comparisons with three popular multimodal multi-objective evolutionary algorithms on three different scenarios of multi-robot task allocation problems are carried out to verify the performance of the proposed algorithm.The experimental test results show that the proposed algorithm can generate sufficient equivalent schemes to improve the availability and robustness of multi-robot collaborative systems in uncertain environments,and also produce the best scheme to improve the overall task execution efficiency of multi-robot collaborative systems.
基金the National Science and Technology Support Program(No.2015BAG20B05)
文摘A new method for a cooperative multi-task allocation problem(CMTAP) is proposed in this paper,taking into account the multi-ship, multi-target, multi-task and multi-constraint characteristics in a multi-ship cooperative driving(MCD) system. On the basis of the general CMTAP model, an MCD task assignment model is established. Furthermore, a genetic ant colony hybrid algorithm(GACHA) is proposed for this model using constraints, including timing constraints, multi-ship collaboration constraints and ship capacity constraints. This algorithm uses a genetic algorithm(GA) based on a task sequence, while the crossover and mutation operators are based on similar tasks. In order to reduce the dependence of the GA on the initial population, an ant colony algorithm(ACA) is used to produce the initial population. In order to meet the environmental constraints of ship navigation, the results of the task allocation and path planning are combined to generate an MCD task planning scheme. The results of a simulated experiment using simulated data show that the proposed method can make the assignment more optimized on the basis of satisfying the task assignment constraints and the ship navigation environment constraints. Moreover, the experimental results using real data also indicate that the proposed method can find the optimal solution rapidly, and thus improve the task allocation efficiency.
文摘在网络状态和任务需求的动态变化下,为减少模型推理时延和计算成本,在软件定义网络(Software Defined Network,SDN)中提出了一种基于模型划分的云边协同推理算法。通过构建复杂度预测器分配任务执行环境,采用深度Q网络(Deep Q-network,DQN)算法对边缘环境中的推理模型进行自适应划分与卸载;以及用SDN技术从全局感知推理任务与网络资源,实现动态网络环境下网络资源的合理分配。试验结果表明:SDN中基于模型划分的云边协同推理算法具有良好的收敛能力,在动态环境中具有较好的鲁棒性。与现有的推理算法相比,该算法能够在合理分配计算资源的前提下,满足协同推理低时延的目标要求。
