Improving the Performance of the Acoustic Emission Method in the Condition Monitoring Process by Optimizing the Placement of Sensors in Oil Refinery Tanks

Abstract

In maintenance, state or condition monitoring, the set of actions describes the state of the machine and its changes over time, based on parameters such as vibrations, sound, performance, lubrication, and temperature. Monitoring the condition of a system requires an optimal network of different sensors to prevent faults occurring in each of the components at an early stage and prevent more severe failures. One of the new methods for monitoring the status is acoustic emission method, which is developed in this paper. The main objective of the present work is to optimize the position of the sensors in monitoring the situation with the method of acoustic emission. The maximum determinant of FISHER Information Matrix (FIM) is the optimal position of the sensors. For this purpose the genetic algorithm is used to find the maximum determinant of Fisher's data matrix for optimal sensor positioning. A case study was conducted on one of the storage tanks of refinery. The design parameters for the positioning of the sensors are respectively the angles relative to the reservoir source and the sensor height relative to the reservoir surface. Finally, the optimal positioning of the sensors is based on the proposed algorithm.