The most important parameters which control the electrolytic process are the concentrations of zinc and sulfuric acid in the electrolyte. An expert control strategy for determining and tracking the optimal concentrati...The most important parameters which control the electrolytic process are the concentrations of zinc and sulfuric acid in the electrolyte. An expert control strategy for determining and tracking the optimal concentrations was proposed, which uses neural networks, rule models and a single loop control scheme. First, the process was described and the strategy that features an expert controller and three single loop controllers was explained. Next, neural networks and rule models were constructed based on statistical data and empirical knowledge on the process. Then, the expert controller for determining the optimal concentrations was designed through a combination of the neural networks and rule models. The three single loop controllers used the PI algorithm to track the optimal concentrations. Finally, the implementation of the proposed strategy were presented. The run results show that the strategy provides not only high purity metallic zinc, but also significant economic benefits.展开更多
The control network is an important developmental orientation in the remote control system. As the control network and information network are comparatively alike in the framework and technology, we can build a contro...The control network is an important developmental orientation in the remote control system. As the control network and information network are comparatively alike in the framework and technology, we can build a control network which is similar to the common information network. In the era when the information network is becoming increasingly mature, it is a royal road to construct or rebuild a control information network in the development of the control network by relying on the achievements made in the information network or current information resources. This paper expounds the construction idea of the control information network, gives the idiographic realization method and then researches into the real-time problem encountered in the control information network, and presents a three-closed-loop control system based on virtualized reality. The feasibility of the idea is validated via experiments and simulations separately.展开更多
Elman networks' dynamical modeling capability is discussed in this paper firstly.According to Elman networks' unique structure,a weight training algorithm is designed and a nonlinear adaptive controller is con...Elman networks' dynamical modeling capability is discussed in this paper firstly.According to Elman networks' unique structure,a weight training algorithm is designed and a nonlinear adaptive controller is constructed.Without the PE presumption,neural networks controller's closed loop properties are studied and the whole Elman networks' passivity is demonstrated.展开更多
当前,知识定义网络赋能AI技术发展,算力网络提供AI所需算力资源,二者逐渐趋于融合,形成了知识定义算力网络(Knowledge Defined Computing Networking,KDCN)。KDCN赋能发展了诸多新型网络应用,如元宇宙、AR/VR、东数西算等,这些新型应用...当前,知识定义网络赋能AI技术发展,算力网络提供AI所需算力资源,二者逐渐趋于融合,形成了知识定义算力网络(Knowledge Defined Computing Networking,KDCN)。KDCN赋能发展了诸多新型网络应用,如元宇宙、AR/VR、东数西算等,这些新型应用对算力资源和网络资源有极大的需求,被称为重击流(Heavy Hitter,HH)。HH流的存在严重加剧了KDCN网络的拥塞情况。针对这一挑战,提出了一种智能流量调度机制,旨在通过深度Q神经网络来解决KDCN中的拥塞问题。相较于离线训练过程,通过流量数据检测与采集、在模型训练和拥塞流调决策之间建立实时闭环,来实现深度Q神经网络模型的在线训练。基于该闭环控制,智能流调模型通过不断学习可以实现持续演化,并用于提供实时决策。实验结果表明,该算法在资源利用率、吞吐量、平均丢包率等方面优于现有方法。展开更多
文摘The most important parameters which control the electrolytic process are the concentrations of zinc and sulfuric acid in the electrolyte. An expert control strategy for determining and tracking the optimal concentrations was proposed, which uses neural networks, rule models and a single loop control scheme. First, the process was described and the strategy that features an expert controller and three single loop controllers was explained. Next, neural networks and rule models were constructed based on statistical data and empirical knowledge on the process. Then, the expert controller for determining the optimal concentrations was designed through a combination of the neural networks and rule models. The three single loop controllers used the PI algorithm to track the optimal concentrations. Finally, the implementation of the proposed strategy were presented. The run results show that the strategy provides not only high purity metallic zinc, but also significant economic benefits.
文摘The control network is an important developmental orientation in the remote control system. As the control network and information network are comparatively alike in the framework and technology, we can build a control network which is similar to the common information network. In the era when the information network is becoming increasingly mature, it is a royal road to construct or rebuild a control information network in the development of the control network by relying on the achievements made in the information network or current information resources. This paper expounds the construction idea of the control information network, gives the idiographic realization method and then researches into the real-time problem encountered in the control information network, and presents a three-closed-loop control system based on virtualized reality. The feasibility of the idea is validated via experiments and simulations separately.
基金This research was supported by the National863Project Foundation(863- 51 1 - 945- 0 1 0 ),Tianjin Nat-ural Science Foundation
文摘Elman networks' dynamical modeling capability is discussed in this paper firstly.According to Elman networks' unique structure,a weight training algorithm is designed and a nonlinear adaptive controller is constructed.Without the PE presumption,neural networks controller's closed loop properties are studied and the whole Elman networks' passivity is demonstrated.
文摘当前,知识定义网络赋能AI技术发展,算力网络提供AI所需算力资源,二者逐渐趋于融合,形成了知识定义算力网络(Knowledge Defined Computing Networking,KDCN)。KDCN赋能发展了诸多新型网络应用,如元宇宙、AR/VR、东数西算等,这些新型应用对算力资源和网络资源有极大的需求,被称为重击流(Heavy Hitter,HH)。HH流的存在严重加剧了KDCN网络的拥塞情况。针对这一挑战,提出了一种智能流量调度机制,旨在通过深度Q神经网络来解决KDCN中的拥塞问题。相较于离线训练过程,通过流量数据检测与采集、在模型训练和拥塞流调决策之间建立实时闭环,来实现深度Q神经网络模型的在线训练。基于该闭环控制,智能流调模型通过不断学习可以实现持续演化,并用于提供实时决策。实验结果表明,该算法在资源利用率、吞吐量、平均丢包率等方面优于现有方法。
