Measurement of gas volume fraction in gas-liquid two-phase flow using arrayed fiber-optic probes combined with the PSO-BP-AdaBoost algorithm

•Proposed a strategy for gas fraction measurement using arrayed fiber-optic probes and neural networks.•Optimizes the structure of arrayed fiber-optic probes to enhance gas phase detection performance.•Constructs a gas phase volume fraction prediction model via the PSO-BP-AdaBoost algorithm.•PSO-BP-...

Celý popis

Uložené v:
Podrobná bibliografia
Vydané v:Optical fiber technology Ročník 93; s. 104264
Hlavní autori: Xing, Wenju, Gao, Hong, Qiao, Xueguang
Médium: Journal Article
Jazyk:English
Vydavateľské údaje: Elsevier Inc 01.09.2025
Predmet:
Adaptive boosting algorithm
Arrayed fiber-optic probes
Backpropagation neural network
Gas volume fraction
Particle swarm optimization algorithm
Adaptive boosting algorithm
Backpropagation neural network
Gas volume fraction
Arrayed fiber-optic probes
Particle swarm optimization algorithm
ISSN:1068-5200
On-line prístup:Získať plný text
Tagy: Pridať tag
Žiadne tagy, Buďte prvý, kto otaguje tento záznam!
Popis
Shrnutí:•Proposed a strategy for gas fraction measurement using arrayed fiber-optic probes and neural networks.•Optimizes the structure of arrayed fiber-optic probes to enhance gas phase detection performance.•Constructs a gas phase volume fraction prediction model via the PSO-BP-AdaBoost algorithm.•PSO-BP-AdaBoost outperforms SVM, BP, and PSO-BP in accuracy, with a max relative error of 0.14%. For the measurement of gas volume fraction in natural gas wells, a strategy based on the fusion of arrayed fiber-optic probes (AFOP) and artificial intelligence algorithms is proposed to enhance the precision and efficiency of gas volume fraction monitoring. As a key front-end component for obtaining gas phase information, AFOP determines the optimal structure by analyzing its performance metrics in bubble capture and its interference with fluid flow. A back-end gas volume fraction prediction model was constructed using a machine learning algorithm. The model first uses a particle swarm optimization (PSO) algorithm to enhance the backpropagation (BP) neural network as a weak predictor and then integrates multiple weak predictors through the adaptive boosting (AdaBoost) algorithm to create a strong predictor. The experimental results show that compared with the support vector machine (SVM), BP neural network, and PSO-BP neural network, the PSO-BP-AdaBoost algorithm has advantages in prediction precision, with a maximum relative error of only 0.14 %, providing a more effective solution for research and application in related fields.
ISSN:1068-5200
DOI:10.1016/j.yofte.2025.104264