Application of Artificial Neural Network-Particle Swarm Optimization Algorithm for Prediction of Asphaltene Precipitation During Gas Injection Process and Comparison With Gaussian Process Algorithm

Abstract

Asphaltene precipitation is a major problem in the oil production and transportation of oil. Changes in pressure, temperature, and composition of oil can lead to asphaltene precipitation. In the case of gas injection into oil reservoirs, the injected gas causes a change in oil composition and may lead to asphaltene precipitation. Accurate determination and prediction of the precipitated amount are vital, for this purpose there are several approaches such as experimental method, scaling equation, thermodynamics models, and neural network as the most recent ones. In this paper, we propose a new artificial neural network (ANN) optimized by particle swarm optimization (PSO) to predict the amount of asphaltene precipitation. This is conducted during the process of gas injection into oil reservoirs for enhanced oil recovery purposes. In the developed models, (1) oil composition, (2) temperature, (3) pressure, (4) oil specific gravity, (5) solvent mole percent, (6) solvent molecular weight, and (7) asphaltene content are considered as input parameters to the neural network. The weight of asphaltene and asphaltene content are considered as input parameters to the neural network and the weight of asphaltene precipitation as an output parameter. A comparison between the results of the proposed new model with Gaussian Process algorithm and previous research shows that the predictive model is more accurate.

Optimization of Wastewater Anaerobic Digestion Using Mechanistic and Meta-heuristic Methods: Current Limitations and Future Opportunities

“…where ω i Inertia weight i max Maximum steps of iteration ω max Maximum value of inertia weight ω min Minimum value of inertia weight Based on empirical studies, acceptable values of ω min = 0.4 while ω max = 0.9 [93].…”

“…Since ANN is heavily dependent on training process and accuracy of model developed is mainly caused by lack in training process, PSO is suggested for weight training as an alternative to use of fuzzy logic. The other reason which contributes to the favoritism of PSO to be coupled with ANN is to acquire faster convergence so that speed of process can be heightened especially for multilayer feedforward network [93]. Initially, PSO will be used to determine the best solution inside the Bsearch^space [91].…”

“…This is due to the fact that high process stability and great reactor performance are observed when fuzzy neural network is applied to the anaerobic system. For future work, further optimization and fine tuning of fuzzy neural network can be achieved such as introduction of adaptive controllers which will be described in detailed in BProcess Control of Anaerobic Digester.^Examples of hybrid PSO with ANN can be found in works presented by Juang and Manshad [91,93] which shows a reduction in absolute relative errors as compared to the use of other models. Other hybrid meta-heuristics in anaerobic digestion can be found in works presented by Tay and Zhang [97], Abu Qdais et al [98], and Rajagopal and Radha [99].…”

“…There are several reasons that support the Fig. 4 Summary of the hybrid ANN and PSO process [93] use of these controllers. Among them are their long history of proven operation, well understood by operational staffs, and have automatic and manual switching, set point tracking, and emergency manual modes.…”

Comparison of supervised and unsupervised approaches for mudstone lithofacies classification: Case studies from the Bakken and Mahantango-Marcellus Shale, USA

“…Schematic architecture of an ANN algorithm where input layer corresponds to different well logs and output refers to different lithofacies (modified afterManshad et al, 2015) …”