Title: Prediction of shale gas horizontal well production using particle swarm optimisation-based BP neural network

Authors: Qi Chen; Wei Wang

Addresses: School of Petroleum Engineering, Guangdong University of Petrochemical Technology, Guangdong Maoming 525000, China ' School of Petroleum Engineering, Guangdong University of Petrochemical Technology, Guangdong Maoming 525000, China

Abstract: The production prediction of shale gas horizontal wells is a critical task. Traditional empirical formulas and mathematical analytical methods have significant errors in predicting shale gas production capacity. To address this issue, we propose a method based on particle swarm optimisation (PSO) to optimise a backpropagation (BP) neural network for shale gas production prediction. Through extensive research, it has been demonstrated that the factors affecting shale gas horizontal well production primarily consist of geological and engineering factors. We employ the grey relational analysis (GRA) method to analyse the main influencing factors on well production and select relevant factors as parameters. The experimental results demonstrate that the utilisation of algorithm-optimised backpropagation (BP) neural networks for predicting the production capacity of hydraulic fracturing wells in actual shale gas reservoirs is more accurate. [Received: August 30, 2023; Accepted: November 21, 2023]

Keywords: particle swarm optimisation; PSO; BP neural network; shale gas; grey correlation method.

DOI: 10.1504/IJOGCT.2024.142105

International Journal of Oil, Gas and Coal Technology, 2024 Vol.36 No.4, pp.449 - 460

Received: 28 Aug 2023
Accepted: 21 Nov 2023
Published online: 07 Oct 2024 *

Full-text access for editors Full-text access for subscribers Purchase this article Comment on this article