MINIMIZING NUMBER OF SENSORS IN WIRELESS SENSOR NETWORKS FOR STRUCTURE HEALTH MONITORING SYSTEMS
MetadataShow full item record
Nowadays, wireless sensor networks (WSNs) are considered an essential candidate to apply structural health monitoring (SHM). An important problem in this area is sensor placement optimization. In many research works, solving this problem focuses only on the network properties and requirements such as energy consumption, network coverage, …etc., without considering the civil engineering requirements. However, there are other research works that consider network and civil requirements while optimizing the sensor placement. Unfortunately, although minimizing the number of sensors is important, it has never been addressed. This could be noticed from the limited literature found that addresses this problem while considering both the civil and the network requirements. As a result, in this thesis we study the problem of minimizing the number of sensors for SHM in WSNs. The idea behind this research is to reduce the network size, which can solve some problems such as the scalability, installation time and cost. Our contribution in this work is not limited to the mathematical model of the mentioned problem, but will extend to solve the problem using different methods: the exhaustive search, genetic algorithm (GA), and a heuristic algorithm that applies the binary search. The problem is then solved for different number of sensors as well as different placements in many conducted experiments. Finally, the time complexity is evaluated to compare between all the applied methods. The obtained results showed that minimizing the number of sensors becomes more significant with big structures. Furthermore, the binary search algorithm is the best to use to solve the problem for small buildings. But, For larger buildings, there is a trade-off between the performance, and time complexity, where binary search gives optimal solution, but genetic algorithm gives better time execution.
- Computer Science & Engineering [34 items ]