Affiliations: [a] School of Mechatronic Engineering, Southwest Petroleum University, Chengdu, Sichuan, China
| [b] Department of Geology, Moscow Lomonosov State University, Russia
Correspondence:
[*]
Corresponding author. Haibo Liang, School of Mechatronic Engineering, Southwest Petroleum University, 8 XinDu Avenue of XinDu District, Chengdu, Sichuan, China. Tel.: +86 18681253466; E-mail: [email protected].
Abstract: To studies and optimizes the fault diagnosis method of comprehensive logging unit pressure sensor, and studies the temperature drift and temperature compensation theory of pressure sensor, establishes fault diagnosis model, and uses artificial intelligence algorithm based on machine learning to apply the temperature drift fault of piezo resistive silicon pressure sensor. The improved particle swarm optimization RBF neural network algorithm has been used to establish the temperature compensation model. By comparing with RBF neural network and particle swarm optimization algorithm, the advantages and disadvantages are compared. The results show that the improved particle swarm optimization RBF neural network algorithm is efficient. The online real-time monitoring of the mud logging sensor network fault has achieved, and the accuracy of the sensor temperature compensation has improved, thereby improving the stability and accuracy of the sensor of the comprehensive logging unit. The simulation results show that the improved particle swarm optimization RBF neural network algorithm has a good temperature compensation effect on the piezoresistive silicon pressure sensor.
