基于人工神经元方式构建的信息物理融合系统  

Cyber Physical System Based on Artificial Neural Way

作　　者：马斌[1] 李楠楠[1] 王永会[1] 邱海斌[2] 张万江[1] 付佳新[1] 

机构地区：[1]沈阳建筑大学信息与控制工程学院,辽宁沈阳110168 [2]沈阳广播电视大学人事处,辽宁沈阳110003

出　　处：《沈阳建筑大学学报（自然科学版）》2012年第2期375-379,共5页Journal of Shenyang Jianzhu University：Natural Science

基　　金：国家科技支撑计划项目(2008BAJ08B08-05);住房和城乡建设部科技项目(2010-K9-49)

摘　　要：目的针对智能建筑室内环境监测,设计信息物理融合系统,使其系统结构具有动态递归形式.方法基于人工神经网络理论构建信息物理融合系统,采用中间件形式进行结构设计,神经网络逼近最优控制率,传播误差设计Kalman滤波器,利用专家经验调整神经网络权值.并以智能建筑室内环境感控为例,说明构建信息物理融合系统对智能建筑技术发展的必然趋势.结果构建的信息物理融合系统是在环境感知的基础上,深度融合计算、通信和控制,可以实现实时感控网络化物理设备.结论人工神经元方式构建的信息物理融合系统在感知的同时进行控制,适用于智能建筑环境感控系统.The paper aims to monitor for intelligent building environment, and design cyber physical system architecture with dynamic recursive form. Based on artificial neural network, we built the cyber physical systems, designed the software structure by the middleware, used neural network approach to optimize the control rate. In addition, we propagation errors by Kalman filter, adjust neural network weights by the expert experience. An intelligent sense of control of environment is built to illustrate that building cyber physical systems of intelligent building technology will be a development trend. The results based on artificial neural approach to build the cyber physical system is perceived in the environment,based on the depth of integration of computing, communication and control ,can achieve real-time infection control network physical device. Building artificial neural in the cyber physical systems adapts to built environment for intelligent control systems.

关 键 词：中间件 信息物理融合 神经网络 智能建筑 

分 类 号：TU318[建筑科学—结构工程]