A WSN Vibration Sensor Nodes Optimization Placement Strategy in Nuclear Power Plant Equipment Monitoring Application

Abstract

Vibration monitoring of equipment is vital to ensure reliable operation of nuclear power plant. Wireless sensor network can reduce monitoring cost, expand monitoring scope, and improve monitoring quality, and vibration sensor nodes placement is a key technology to guarantee the accuracy and correctness of vibration data acquisition and further analysis. Current placement researches are unable to meet the requirements of nuclear power plant application. In this paper, we first describe the problem and present its mathematical model, and then we developed a strategy based on genetic algorithm to address this problem. Considering constraints of Modal Information, sensitive nuclear power plant devices, connectivity, security, maintenance convenience, and placement convenience, we compare artificial placement, random placement, and our placement algorithm. Results show that fitness value of artificial placement and random placement is respectively 1.18 and 1.55 times as large as our algorithm's. We also test the impact of cross probability, mutation probability, and population size on our algorithm. Simulation results show that our algorithm is efficient, with run time 16% of enumeration algorithm.