Fault-tolerant sensor integration for micro flow-sensor arrays and networks

Abstract

Distributed micro flow-sensor arrays and networks (DMFSA/N), built from collections of spatially scattered, cooperating intelligent and redundant micro flow-sensor nodes, can improve the accuracy and reliability of system. However, it is unrealistic to expect all the sensor nodes and communication links in the system to function properly all the time. This paper is based on an earlier research, in which a DMFSA/N with a cluster architecture and fault-tolerant time-out Protocol (FTTP) was developed. In this paper, a fault-tolerant sensor integration algorithm (FTSIA) is proposed and evaluated in simulations. Experimental results showed that the FTSIA could always give reliable results even when certain portion of the sensors yielded faulty information. Furthermore, the results from FTSIA were significantly more accurate than the mean of sensor readings (popularly applied in industry) if some of the sensors produced faulty readings. Finally, the application of the proposed FTSIA is illustrated by measuring flow pressure using a pressure sensor array of eight sensors.