摘要
信号重构是计算机系统与被控对象进行信息通信的重要环节,在简要分析Shannon重构、ZOH重构和一阶重构等信号重构方法局限性的基础上,提出了一组基于三点插值的重构算法和被控过程采样数据的一步容错预测算法,并将一步容错预测与三点插值算法相结合,建立了对于被控过程信号异常情况有良好容错能力的一步容错预测二阶重构算法。上述的一步容错预测二阶重构算法不仅结构简单、可用于在线控制,而且算法精度和可靠性明显高于普通的ZOH重构和一阶重构算法。
Signal reconstruction is an important step in communication between the computer controlled system and computer system. Based on the famous Shannon reconstructing theory, some classical algorithms (such as ZOH algorithm and 1-order reconstruction algorithm etc) were analyzed and some shortages of these algorithms were pointed out. And, a new practical 1-step fault-tolerant prediction algorithm was set up. In order to realize this fault-tolerant algorithm, the three-points-interpolation algorithm was combined into the reconstruction process. This new algorithm has the ability to overcome bad influence from outliers and has very simple computing complexity. So, this algorithm can be safely used in on-line control.
出处
《系统仿真学报》
CAS
CSCD
北大核心
2006年第z2期841-843,共3页
Journal of System Simulation
基金
国家自然科学基金(90305007)
中-英联合项目
中国博士后科学基金(0503772)
江苏省博士后科研资助计划(0501004A)
瑞典皇家学院研究基金(SI-210-05483-2005)
关键词
信号重构
容错算法
计算机控制系统
signal reconstruction
fault-tolerant algorithm
computer controlled system