Fault-tolerant sensor integration for micro flow-sensor arrays and networks
Section snippets
Problem background
Conventionally, single sensor systems have been used to save space and manufacturing cost. But they are not suitable for critical applications because a single sensor is inevitably affected by noise, refresh delay and other uncertainty issues and cannot guarantee to deliver accurate information all the time. Development of MEMS (Micro Electro Mechanical Systems) has enabled production of relatively large number of micro sensors at a low cost. Micro flow-sensors are one of the most common MEMS
A cluster architecture with fault-tolerant time-out protocol for DMFSA/N
Since micro flow-sensor nodes are distributed spatially in DMFSA/N, a suitable architecture has to be designed to enable efficient and fault-tolerant communication and integration in the system. A cluster architecture of DMFSA/N was proposed and developed in Liu and Nof (2004). In this structure, DMFSA/N are divided into sensor cluster units (SCUs), each of which consists of a set of intelligent micro flow-sensor nodes and a base station. The intelligent sensor nodes within the same SCU are
Fault-tolerant sensor integration algorithm
Fault-tolerant sensor integration algorithm (FTSIA) refers to the algorithm used to combine information from different sensors in the system and produce reliable results even if some sensors yield faulty information. The algorithm developed in this research follows the ideas originated from (Marzullo, 1990) and later improved by Jayasimha (1996). It deals with the competitive integration of sensor information, in which each sensor ideally measures identical information, but, in reality, is
Case study with actual sensor array
This section illustrates the application of the FTSIA described in Section 3 through measuring flow pressure using an actual pressure sensor array of eight sensors.
Conclusions and future tasks
This paper proposes a fault-tolerant sensor integration algorithm (FTSIA) to fuse signals from multiple individual sensor nodes in DMFSA/N, considering some of them may be faulty. Three experiment cases were simulated to test the reliability of the proposed FTSIA, and the results from FTSIA were compared with the mean of sensor readings. The simulation experiment results showed that when ft was large relative to Ns and when fw > 0, the FTSIA yielded more accurate results than the mean. The
References (18)
- et al.
Prandtl micro flow sensor (PMFS): A novel silicon diaphragm capacitive sensor for flow-velocity measurement
Sensors & Actuators A-Physical
(1998) Sensor array design for wave decomposition in the presence of coupled motion
Journal of Sound & Vibration
(2003)- et al.
Measurement of flow direction and velocity using a micromachined flow sensor
Sensors and Actuators A (Physical)
(2004) - et al.
Electrochemical time of flight flow sensor
Sensors and Actuators, A: Physical
(2002) - et al.
A MEMS multi-sensor chip for gas flow sensing
Sensors and Actuators, A: Physical
(2005) - et al.
Time-out protocol for task allocation in multi-agent systems
Journal of International Manufacturing
(2002) - et al.
E-work: The challenge for next generation ERP systems
Production Planning & Control
(2003) - et al.
A client-server model for integration and collaboration in production testing
International Journal of Production Research
(1998) - et al.
Tracking and imaging humans on heterogeneous infrared sensor array for tactical applications
Proceedings of SPIE – The International Society for Optical Engineering
(2002)
Cited by (3)
A resilience by teaming framework for collaborative supply networks
2015, Computers and Industrial EngineeringCitation Excerpt :FTT implies that, under the right protocols, a set of collaborative agents that are more prone to failure can outperform a single flawless agent. Liu and Nof (2004, 2008) apply the FTT principle to the design of and operation of sensor networks and develop: (1) a sensor network architecture capable of leveraging sensors with low reliability, (2) a protocol for timely supply of sensor reading, and (3) a sensor reading integration algorithm to overcome faulty readings. Experimental results show that the aforementioned are able to (1) overcome faults while ensuring low energy consumption, (2) provide more accurate sensor readings – both for mean value accuracy and reading interval length – than an accurate single sensor, and (3) outperform traditional protocols in both small and large networks.
Resilience by teaming in supply network formation and re-configuration
2015, International Journal of Production EconomicsCitation Excerpt :Then, collaboration among these agents makes the system more robust and tolerant to failure than any one of the agents. Based on the FTT principle, Liu and Nof (2004; 2008) introduce (i) a sensor network architecture capable of leveraging sensors with low reliability, (ii) an operational protocol for timely supply of sensor readings, and (iii) a sensor reading integration algorithm to overcome faulty readings. Jeong and Nof (2009) propose a simulated annealing algorithm to dynamically form sensor clusters in a network in order to minimize energy consumption in message transmission.
Fault tolerant interval fusion for moving vehicle classification in wireless sensor networks
2009, IET Conference Publications