Unmanned aerial vehicle(UAV)-enabled edge computing is emerging as a potential enabler for Artificial Intelligence of Things(AIoT)in the forthcoming sixth-generation(6G)communication networks.With the use of flexible ...Unmanned aerial vehicle(UAV)-enabled edge computing is emerging as a potential enabler for Artificial Intelligence of Things(AIoT)in the forthcoming sixth-generation(6G)communication networks.With the use of flexible UAVs,massive sensing data is gathered and processed promptly without considering geographical locations.Deep neural networks(DNNs)are becoming a driving force to extract valuable information from sensing data.However,the lightweight servers installed on UAVs are not able to meet the extremely high requirements of inference tasks due to the limited battery capacities of UAVs.In this work,we investigate a DNN model placement problem for AIoT applications,where the trained DNN models are selected and placed on UAVs to execute inference tasks locally.It is impractical to obtain future DNN model request profiles and system operation states in UAV-enabled edge computing.The Lyapunov optimization technique is leveraged for the proposed DNN model placement problem.Based on the observed system overview,an advanced online placement(AOP)algorithm is developed to solve the transformed problem in each time slot,which can reduce DNN model transmission delay and disk I/O energy cost simultaneously while keeping the input data queues stable.Finally,extensive simulations are provided to depict the effectiveness of the AOP algorithm.The numerical results demonstrate that the AOP algorithm can reduce 18.14%of the model placement cost and 29.89%of the input data queue backlog on average by comparing it with benchmark algorithms.展开更多
This paper proposes a multi-objective index-based approach to optimally determine the size and location of multi-distributed generators (DG) units in distribution system with different load models. It is shown that lo...This paper proposes a multi-objective index-based approach to optimally determine the size and location of multi-distributed generators (DG) units in distribution system with different load models. It is shown that load models can significantly affect the optimal location and sizing of DG resources in distribution systems. The proposed multi-objective function to be optimized includes a short circuit level parameter to represent the protective device requirements. The proposed function also considers a wide range of technical issues such as active and reactive power losses of the system, the voltage profile, the line loading and the MVA intake by the grid. The optimization technique based on particle swarm optimization (PSO) is introduced. The analysis of continuation power flow to determine the effect of DG units on the most sensitive buses to voltage collapse is carried out. The proposed algorithm is tested using the 38-bus radial system and the IEEE 30-bus meshed system. The results show the effectiveness of the proposed algorithm.展开更多
近年来,大模型推动自然语言处理、机器视觉等众多领域取得前所未有的进展.混合专家(mixture of experts,MoE)凭借在模型参数扩展、计算成本控制和复杂任务处理等方面的独特优势成为大模型的主流架构之一.然而,随着参数规模的持续增长,...近年来,大模型推动自然语言处理、机器视觉等众多领域取得前所未有的进展.混合专家(mixture of experts,MoE)凭借在模型参数扩展、计算成本控制和复杂任务处理等方面的独特优势成为大模型的主流架构之一.然而,随着参数规模的持续增长,系统的执行效率和可扩展能力愈发难以满足需求,亟待解决.系统优化方法是解决这一挑战的有效途径,日益成为研究热点.故综述大模型时代MoE系统优化技术的研究现状,首先介绍MoE大模型的发展现状,并分析其在系统端面临的性能瓶颈;然后从内存占用、通信延迟、计算效率和并行扩展4个系统核心维度对最新的研究进展进行全面梳理和深入分析,并对其中涉及的关键技术、适用场景和待优化方向进行详细对比阐述;最后总结MoE系统优化的研究现状,并展望未来研究方向.展开更多
By taking advantage of the separation characteristics of nonlinear gain and dynamic sector inside a Hammerstein model, a novel pole placement self tuning control scheme for nonlinear Hammerstein system was put forward...By taking advantage of the separation characteristics of nonlinear gain and dynamic sector inside a Hammerstein model, a novel pole placement self tuning control scheme for nonlinear Hammerstein system was put forward based on the linear system pole placement self tuning control algorithm. And the nonlinear Hammerstein system pole placement self tuning control(NL-PP-STC) algorithm was presented in detail. The identi fication ability of its parameter estimation algorithm of NL-PP-STC was analyzed, which was always identi fiable in closed loop. Two particular problems including the selection of poles and the on-line estimation of model parameters, which may be met in applications of NL-PP-STC to real process control, were discussed. The control simulation of a strong nonlinear p H neutralization process was carried out and good control performance was achieved.展开更多
基金supported by the National Science Foundation of China(Grant No.62202118)the Top-Technology Talent Project from Guizhou Education Department(Qianjiao Ji[2022]073)+1 种基金the Natural Science Foundation of Hebei Province(Grant No.F2022203045 and F2022203026)the Central Government Guided Local Science and Technology Development Fund Project(Grant No.226Z0701G).
文摘Unmanned aerial vehicle(UAV)-enabled edge computing is emerging as a potential enabler for Artificial Intelligence of Things(AIoT)in the forthcoming sixth-generation(6G)communication networks.With the use of flexible UAVs,massive sensing data is gathered and processed promptly without considering geographical locations.Deep neural networks(DNNs)are becoming a driving force to extract valuable information from sensing data.However,the lightweight servers installed on UAVs are not able to meet the extremely high requirements of inference tasks due to the limited battery capacities of UAVs.In this work,we investigate a DNN model placement problem for AIoT applications,where the trained DNN models are selected and placed on UAVs to execute inference tasks locally.It is impractical to obtain future DNN model request profiles and system operation states in UAV-enabled edge computing.The Lyapunov optimization technique is leveraged for the proposed DNN model placement problem.Based on the observed system overview,an advanced online placement(AOP)algorithm is developed to solve the transformed problem in each time slot,which can reduce DNN model transmission delay and disk I/O energy cost simultaneously while keeping the input data queues stable.Finally,extensive simulations are provided to depict the effectiveness of the AOP algorithm.The numerical results demonstrate that the AOP algorithm can reduce 18.14%of the model placement cost and 29.89%of the input data queue backlog on average by comparing it with benchmark algorithms.
文摘This paper proposes a multi-objective index-based approach to optimally determine the size and location of multi-distributed generators (DG) units in distribution system with different load models. It is shown that load models can significantly affect the optimal location and sizing of DG resources in distribution systems. The proposed multi-objective function to be optimized includes a short circuit level parameter to represent the protective device requirements. The proposed function also considers a wide range of technical issues such as active and reactive power losses of the system, the voltage profile, the line loading and the MVA intake by the grid. The optimization technique based on particle swarm optimization (PSO) is introduced. The analysis of continuation power flow to determine the effect of DG units on the most sensitive buses to voltage collapse is carried out. The proposed algorithm is tested using the 38-bus radial system and the IEEE 30-bus meshed system. The results show the effectiveness of the proposed algorithm.
文摘近年来,大模型推动自然语言处理、机器视觉等众多领域取得前所未有的进展.混合专家(mixture of experts,MoE)凭借在模型参数扩展、计算成本控制和复杂任务处理等方面的独特优势成为大模型的主流架构之一.然而,随着参数规模的持续增长,系统的执行效率和可扩展能力愈发难以满足需求,亟待解决.系统优化方法是解决这一挑战的有效途径,日益成为研究热点.故综述大模型时代MoE系统优化技术的研究现状,首先介绍MoE大模型的发展现状,并分析其在系统端面临的性能瓶颈;然后从内存占用、通信延迟、计算效率和并行扩展4个系统核心维度对最新的研究进展进行全面梳理和深入分析,并对其中涉及的关键技术、适用场景和待优化方向进行详细对比阐述;最后总结MoE系统优化的研究现状,并展望未来研究方向.
文摘By taking advantage of the separation characteristics of nonlinear gain and dynamic sector inside a Hammerstein model, a novel pole placement self tuning control scheme for nonlinear Hammerstein system was put forward based on the linear system pole placement self tuning control algorithm. And the nonlinear Hammerstein system pole placement self tuning control(NL-PP-STC) algorithm was presented in detail. The identi fication ability of its parameter estimation algorithm of NL-PP-STC was analyzed, which was always identi fiable in closed loop. Two particular problems including the selection of poles and the on-line estimation of model parameters, which may be met in applications of NL-PP-STC to real process control, were discussed. The control simulation of a strong nonlinear p H neutralization process was carried out and good control performance was achieved.
