过程控制系统中多条件约束的多传感器故障检测与诊断

Multi-sensor Fault Detection and Diagnosis Scheme with Multiple Constraints for Process Control System

摘要:

针对工业过程控制系统中的故障具有类型多样、时空独立和非线性等特点,使得检测与诊断效率降低,系统性能下降等问题,提出了一种基于自定义多条件约束的多传感器故障检测与诊断机制；该机制,首先考虑了系统的稳态和时空特征建立了非线性过程控制系统多故障模型,并给出了满足条件判定法则；然后对于系统中的单故障,并发故障和通信故障等类型给出了多条件约束法则及独立特性判断；最后提出了通过自定义多条件约束的多传感器故障检测与诊断机制；实验结果表明,在平均检测概率、稳态特征保持能力和系统功耗等方面明显优于无条件约束的机制,可以显著改善过程控制系统性能。

Abstract:

For industrial process control system fault has diverse, independent and non-linear temporal characteristics, making the detection and diagnosis of reduced efficiency, system performance and other issues, we propose a multi-sensor based on a custom multi-fault detection and diagnosis of constraints mechanisms. This mechanism, first consider the steady-state and temporal characteristics of the system to establish a nonlinear process control system multi-fault model, and gives the condition judgment rule; then for a single fault in the system, concurrent failures and communication failures and other types of shows multi-constraint rules and independent properties judgment; concludes with a multi-sensor fault detection and diagnosis mechanisms by customizing multiple constraints. Experimental results show that the average probability of detection, the ability to maintain steady-state characteristics and system power consumption, significantly better than the unconditional restraint mechanisms can significantly improve the performance of the process control system.