点击次数：

论文类型：期刊论文

发表时间：2014-06-01

发表刊物：INTERNATIONAL JOURNAL OF STRUCTURAL STABILITY AND DYNAMICS

收录刊物：SCIE、EI

卷号：14

期号：5

ISSN号：0219-4554

关键字：Structural health monitoring; wireless sensor network; optimal sensor placement; generalized genetic algorithm; self-adaptive dynamic penalty function

摘要：The optimal placement of wireless sensors is very different from conventional wired sensor placement due to the limited transmission range of the wireless sensors. This constraint on the inter-sensor distance makes the optimization problem difficult to solve with conventional gradient-based methods. In this paper, an improved generalized genetic algorithm (GGA) based on a self-adaptive dynamic penalty function (SADPF) is proposed for the optimal wireless sensor placement (OWSP) in bridge vibration monitoring. The mathematical model of the OWSP problem is established, and it considers both the bridge vibration monitoring requirements and the constraints of the data transmission range in wireless sensor networks (WSNs). SADPF, which can automatically adjust the amount of penalization for constraint violations according to the evolution generation number and the degree of violation, is then developed so that the wireless sensor placement can be optimized using GGA. Subsequently, the GGA is improved by implementing an elite conservation strategy, a worst elimination policy and a dual-structure coding system. Finally, a numerical experiment is presented with a long-span suspension bridge to demonstrate the feasibility and efficiency of the proposed method, and some indispensible discussions are also given. The results indicate that the wireless sensor configurations that are optimized by the improved SADPF-based GGA can simultaneously meet the data transmission demands in a WSN and fulfill the requirements for structural condition assessment. The developed SADPF can minimize the influence of the limited data transmission range on the search process for the OWSP. The improved SADPF-based GGA quickly and robustly converges to the global optimal solution.