Multi-agent system design and evaluation for collaborative wireless sensor network in large structure health monitoring

https://doi.org/10.1016/j.eswa.2009.06.098Get rights and content

Abstract

Much attention has been focused on the research of structural health monitoring (SHM), since it could increase the safety and reduce the maintenance costs of engineering structures. In recent years, wireless sensor network (WSN) has been explored for adoption to improve the centralized cable-based SHM system performances. This paper presents a multi-agent design method and system evaluation for wireless sensor network based structural health monitoring to validate the efficiency of the multi-agent technology. Through the cooperation of six different agents for SHM applications, the distributed wireless sensor network can automatically allocate SHM tasks, self-organize the sensor network and aggregate different sensor information. In the evaluation work, the strain gauge and PZT sensors are used to monitor strain distribution change and joint failure of an experimental aluminum plate structure. A dedicated sensor network platform including the wireless strain node, wireless PZT node and wireless USB station is designed for the evaluation system. Based on the hardware platform, the multi-agents software architecture is defined. The multi-agent monitoring principle and implementation in the validation work for two typical kinds of structure states are presented. This paper shows the efficiency of the multi-agent technology for WSN based the SHM applications on the large aircraft structures.

Introduction

The performance of the in-service aerospace structures can be affected by degradation resulting from exposure to harsh flight environment conditions or damages resulting from external conditions, such as impact, loading, operator abuse or neglect. In order to improve the safety level of the aerospace structures, structural health monitoring (SHM) is researched for devoted to predict the onset of damage and deterioration in aircraft structural condition with the observation of a system over time using periodically sampled dynamic response measurements from an array of sensors. Traditionally, the cable-based SHM systems for aircraft structures might involve large number of wires employed for communication among sensors and centralized data acquisition systems. If the SHM system designed for a large structure is comprised of a large number of sensors and address stringent real-time operations, the system can become overburdened with computational tasks. In response to the cumbersome wires and performance shortcomings of centralized cable-based SHM systems, wireless sensor network (WSN) has been explored for adoption in recent years (Caffrey et al., 2004, Lynch et al., 2003, Straser and Kiremidjian, 1998, Xu et al., 2004).

For SHM system applied on large scale structure, different kinds of density sensor networks are required to monitor different structure parameters, such as stress, stain, displacement, acoustic, pressure, temperature and etc. These sensors with different theories and functions might be connected to the wireless sensor nodes for data acquisition and processing. Many benefits can be gained from wireless SHM system, such as local computational ability, low cost deployment, and wireless networking functionality. However, the information obtained by each sensor node is limited, so does the local signal processing ability of each sensor node. Besides, real large scale aircraft structures are complicated to be estimated. Hence, when implementing a wireless SHM system for practical big aircraft structures, the challenge might includes how to integrate dedicated wireless SHM sensor nodes; how to coordinate and manage the large dense wireless SHM network, since the SHM systems for practical aircraft structures need multi-point, heterogeneous, and synchronous operations.

Multi-agent technology over the past few years has come to be perceived as crucial technology not only for effectively exploiting the increasing availability of diverse, heterogeneous and distributed on-line information sources, but also as a framework for building large, complex and robust distributed information processing systems which exploit the efficiencies of organized behavior. Given the general benefits of multi-agents, scholars have explored the possibility for sensor network applications (Helvik and Wittner, 2001, Kumar et al., 2002, Qi et al., 2001). A design method for MAS based SHM system has been presented by the authors (Yuan, Lai, & Zhao, 2006). This paper presents a multi-agent design method and system evaluation for wireless sensor network based structural health monitoring to validate the efficiency of the multi-agent technology. The collaborative multi-agent wireless sensor network architecture including the physical and software layers is presented. Six cooperation agents for SHM applications are defined based on the clustered wireless sensor network which could automatically allocate SHM tasks, self-organize the sensor network and aggregate different sensor information. In the evaluation work, the strain gauge and PZT sensors are used to monitor strain distribution change and joint failure of an aircraft aluminum plate structure. A dedicated sensor network platform including the wireless strain node, wireless PZT node and wireless USB station is designed for the evaluation system. Based on the hardware platform, the multi-agent software implementations are introduced in the validation work for two typical kinds of structure states. The evaluation system and experimental result shows the efficiency of the multi-agent technology for WSN based SHM applications on large aircraft structures.

Section snippets

Collaborative MA-WSNs architecture based SHM

To apply the multi-agent technology to the distributed WSN based SHM applications, a multi-agent architecture for sensor networks that can support collaborative SHM task allocation, network self-organization and sensor data aggregation is described from both physical and software architecture perspectives.

Multi-agent WSN architecture evaluation for SHM

In order to validate the efficiency of the multi-agent WSN architecture for SHM, an aircraft plate structure as the typical engineering structure is adopted. Two typical structure states which may be monitored by sensor arrays and indicate structural damages are researched such as joint failure and strain distribution change.

Conclusions

The evaluation work shows that multi-agent WSN based SHM could gain the following advantages. The collaborative multi-agent wireless sensor network architecture could automatically allocate SHM tasks by the mobile agent (SAA). According to different SHM task trigger mode, such as periodic triggered tasks and user triggered tasks, static and dynamical itinerary planning might be used for the design of the itinerary of SAA. Clustered sensor network can be changed and self-organized by SAA during

Acknowledgements

This work is supported by Natural Science Foundation of China (Grant Nos. 60772072 and 50830201), National High-Tech Research and Development Plan (863 Plan). (Grant No. 2007AA03Z117).

References (10)

  • Caffrey, J., Govindan, R., Johnson, E., Krishnamachari, B., Masri, S., & Sukhatme, G., et al. (2004). Networked sensing...
  • Helvik, B. E., & Wittner, O. (2001). Using the cross-entropy method to guide/govern mobile agent’s path finding in...
  • Hill, J., (2003). System architecture for wireless sensor networks. Ph.D. thesis, UC...
  • S. Kumar et al.

    Collaborative signal and information processing in micro-sensor networks

    IEEE Signal Processing Magazine

    (2002)
  • Lynch, J. P., Sundararajan, A., Law, K., Kiremidjian, A., Carryer, E., & Sohn, H. (2003). Field validation of a...
There are more references available in the full text version of this article.

Cited by (0)

View full text