基于动态神经网络的时延自动化控制数学建模研究  

Research on Mathematical Modeling of Time Delay Automation Control Based on Dynamic Neural Network

作　　者：王于琴 田峰 WANG Yuqin;TIAN Feng(Chongqing Vocational and Technical University of Mechatronics,Chongqing 402760,China)

机构地区：[1]重庆机电职业技术大学,重庆402760

出　　处：《自动化与仪器仪表》2022年第5期79-83,共5页Automation & Instrumentation

基　　金：新时代职业技术本科课程新形态数学教材建设的范式研究与实践(CMC20210108)。

摘　　要：为构建控制误差更低、上行负载降低效果更好的时延自动化控制数学模型,借鉴现有研究成果,设计一种基于动态神经网络的时延自动化控制数学建模方法。构建移动带宽动态分配模型,统一调度PON资源与无线资源。构建的移动带宽动态分配模型由基带处理单元、无输出变压器等构成。在通信过程中分析时延不确定因素,通过概率图模型化简时延不确定因素所构成的概率模型,使概率模型的后续学习更快,时延自动化控制的搜索速率更高。基于动态神经网络设计一种在线学习算法,使概率模型化简结果能够在线学习,通过概率图在线学习实现时延自动化控制,对该过程实施数学建模。仿真LTE网络和无源光纤网络通信参数,测试设计方法的控制性能。测试结果为设计方法在LTE网络规模增大的情况下能够保持较低的控制误差,能够实现通信系统上行负载的降低,降低系统的通信负担。说明设计方法能够很好地服务于各种通信系统。In order to construct a time-delay automation control mathematical model with lower control error and better uplink load reduction effect,based on the existing research results,a mathematical modeling method of time-delay automation control based on dynamic neural network is designed.Build a dynamic allocation model of mobile bandwidth,and uniformly schedule PON resources and wireless resources.The constructed mobile bandwidth dynamic allocation model is composed of a baseband processing unit and no output transformer.Analyze the delay uncertain factors in the communication process,and simplify the probability model formed by the delay uncertain factors through the probability graph model,so that the subsequent learning of the probability model is faster,and the search rate of the delay automation control is higher.Based on the dynamic neural network,an online learning algorithm is designed to enable online learning of the simplified results of the probabilistic model,realize the automatic control of time delay through online learning of the probability diagram,and implement mathematical modeling of the process.Simulate LTE network and passive optical fiber network communication parameters,and test the control performance of the design method.The test result is that the design method can maintain a low control error when the LTE network scale is increased,can reduce the uplink load of the communication system,and reduce the communication burden of the system.It shows that the design method can serve various communication systems well.

关 键 词：动态神经网络 时延自动化控制 概率图模型 数学建模 时延不确定因素 在线学习算法 

分 类 号：TH113.1[机械工程—机械设计及理论]