Study on Sensor Network Optimization for MS/AE Monitoring System Using Fisher Information and Improved Encoding Framework

The current source localization is severely challenged by the geometric layout of the sensor network. To eliminate the influence of sensor network on the localization accuracy, this article proposes a novel optimization method of sensor networks for microseismic/acoustic emission (MS/AE) monitoring systems using a Fisher information and an improved encoding framework. First, the range difference (RD) equation of MS/AE source is constructed for the system with known wave velocity. Second, the optimization function of sensor network layout is constructed using the Fisher information and probability integral. Third, the encoding sensor domains, combined with improved genetic manipulations, are adopted to solve the optimal sensor network. The feasibility and effectiveness of the proposed method are verified by laboratory AE experiments. The results demonstrate that the optimal sensor network can be efficiently identified by the proposed method, and the relative positioning accuracy of four-sensor optimal network reaches 3.7%. Simulation experiments are further conducted to investigate the localization performance of the common sensor networks and the geometric configuration of the optimal network. In addition, general rules for sensor network layouts in systems with known velocities are given to guide the placement of sensors in daily monitoring systems.