摘要
研究一类测量信号具有量化误差的网络化系统的故障检测问题.系统的每一路测量输出采用一个独立的静态对数量化器,通过扇形界方法描述对数量化器带来的量化误差,将系统的参数不确定转化为范数不确定,利用线性矩阵不等式方法设计H-/H∞故障检测系统.所设计的故障检测系统的残差是渐近稳定的,对外界干扰等不确定输入具有鲁棒性,同时对故障信号具有较高的灵敏度.最后,通过一个数值例子指出测量信号量化带来的量化误差降低了故障检测系统的鲁棒性,忽略测量信号的量化误差可能导致故障检测系统发生误报等不良现象.数值仿真结果表明,设计的故障检测系统是可靠有效的.
A fault detection problem of networked control systems with quantized measurements is discussed. First,a separate and static logarithmic quantizer for each output is modeled. The sector bound approach is used to describe the quantization error. Then the quantized fault detection problem is transformed into the fault detection problem for an uncertain system,and an H /H∞ fault detection system is designed in terms of linear matrix inequalities. The residuals of this fault detection system are asymptotically stable,which are robust against uncertain disturbances and sensitive to fault. Finally,a numerical example is provided to verify that the quantization error decreases the fault detection system's robustness and accurateness. Furthermore,the simulation result demonstrates the validity of the designed fault detection system.
出处
《东南大学学报(自然科学版)》
EI
CAS
CSCD
北大核心
2010年第S1期255-260,共6页
Journal of Southeast University:Natural Science Edition
基金
国家自然科学基金资助项目(60874059
60604027)
福建省自然科学基金资助项目(2009J01279)
关键词
网络控制系统
量化
故障检测
鲁棒性
灵敏度
networked control systems
quantization
fault detection
robust
sensitivity