线性测量网络的优化设计  

Optimal design of sensor networks for linear processes

作　　者：周凌柯 苏宏业[1] 褚健[1] 

机构地区：[1]浙江大学先进控制研究所

出　　处：《浙江大学学报（工学版）》2005年第6期853-857,共5页Journal of Zhejiang University：Engineering Science

基　　金：国家杰出青年科学基金资助项目(60025308);国家"十五"科技攻关资助项目(2001BA204B01).

摘　　要：针对传统的线性测量网络设计中没有给出具体的整数线性规划(ILP)模型问题,提出一种传感器优化配置模型.通过分析测量网络回路,推出了变量冗余度的数学表达式,给出变量冗余度的上界来避免测量网络变量的不可行冗余度约束,并结合图论方法和ILP方法,建立了以费用最小为目标的最小测量网络和满足指定变量冗余度要求的冗余测量网络的传感器ILP模型,并对一个冗余测量网络和一个非冗余的蒸汽测量网络进行测量网络设计.仿真结果表明,利用该模型能够得到满足冗余度要求及仪表配置费用最低的优化测量网络.To resolve the problem of the traditional linear sensor networks design not having detailed integer linear programming (ILP) model, an optimal redundant sensor network model was proposed. The mathematical representation of redundancy degree was introduced based on analysis of the cycles of sensor networks, and the bound of redundancy degree was analyzed to avoid the infeasible degree of redundancy given. By combining theoretical graph concepts and ILP methods, the optimal design featuring minimum cost of nonredundant sensor networks, as well as redundant sensor networks that satisfy constraints related to the redundancy degree of key variables were examined. Applications to a redundant sensor network and a nonredundant steam-metering sensor network show that the optimal sensor network can ensure the redundancy degree of variables and realize the minimum sensor location investment by using this model.

关 键 词：数据校正 测量网络设计 图论 可观性 冗余性 

分 类 号：TP202[自动化与计算机技术—检测技术与自动化装置]