Injection and Production Optimization for Tengiz Platform Sour Gas Injection by Reactive and Proactive Conformance Completions Control

2021 ◽  
Author(s):  
Shusei Tanaka ◽  
Matthieu Rousset ◽  
Yousef Ghomian ◽  
Aizada Azhigaliyeva ◽  
Chingiz Bopiyev ◽  
...  
Keyword(s):  
Machine Learning ◽  
Genetic Algorithm ◽  
Gas Injection ◽  
Hybrid Approach ◽  
Optimization Techniques ◽  
Control Parameters ◽  
Objective Functions ◽  
Conformance Control ◽  
Model Based ◽  
Sour Gas

Abstract Sour gas injection operation has been implemented in Tengiz since 2008 and will be expanded as part of a future growth project. Due to limited gas handling capacity, producing wells at high GOR has been a challenge, resulting in potential well shutdowns. The objective of this study was to establish an efficient optimization workflow to improve vertical/areal sweep, thereby maximizing recovery under operation constraints. This will be enabled through conformance control completions that have been installed in many production/injection wells. A Dual-Porosity and Dual-Permeability (DPDK) compositional simulation model with advanced Field Management (FM) logic was used to perform the study. Vertical conformance control was implemented in the model enabling completion control of 4 compartments per well. A model-based optimization workflow was defined to maximize recovery. Objective functions considered were incremental recovery 1) after 5 years, and 2) at the end of concession. Control parameters considered for optimization are 1) injection allocation rate, 2) production allocation rate, 3) vertical completion compartments for injectors and producers. A combination of different optimization techniques e.g., Genetic Algorithm and Machine-Learning sampling method were utilized in an iterative manner. It was quickly realized that due to the number of mixed categorical and continuous control parameters and non-linearity in simulation response, the optimization problem became almost infeasible. In addition, the problem also became more complex with multiple time-varying operational constraints. Parameterization of the control variables, such as schedule and/or FM rules optimization were revisited. One observation from this study was that a hybrid approach of considering schedule-based optimization was the best way to maximize short term objectives while rule-based FM optimization was the best alternative for long term objective function improvement. This hybrid approach helped to improve practicality of applying optimization results into field operational guidelines. Several optimization techniques were tested for the study using both conceptual and full-field Tengiz models, realizing the utility of some techniques that could help in many field control parameters. However, all these optimization techniques required more than 2000 simulation runs to achieve optimal results, which was not practical for the study due to constraints in computational timing. It was observed that limiting control parameters to around 50 helped to achieve optimal results for the objective functions by conducting 500 simulation runs. These limited number of parameters were selected from flow diagnostics and heavy-hitter analyses from the pool of original 800+ control parameters. The novelty of this study includes three folds: 1) The model-based optimization outcome obtained in this study has been implemented in the field operations with observation of increased recovery 2) the hybrid optimization of both schedule and operation rule provided practicality in terms of optimization performance as well as application to the field operation 3) provides lessons learned from the application of optimization techniques ranging from conventional Genetic Algorithm to Machine-Learning supported technique.

Simulation of English text recognition model based on ant colony algorithm and genetic algorithm

2021 ◽  
pp. 1-12
Author(s):  
Fei Long
Keyword(s):  
Machine Learning ◽  
Genetic Algorithm ◽  
Ant Colony Algorithm ◽  
Ant Colony ◽  
Text Recognition ◽  
English Text ◽  
Recognition Model ◽  
Central Node ◽  
Model Based ◽  
Characteristic Points

The difficulty of English text recognition lies in fuzzy image text classification and part-of-speech classification. Traditional models have a high error rate in English text recognition. In order to improve the effect of English text recognition, guided by machine learning ideas, this paper combines ant colony algorithm and genetic algorithm to construct an English text recognition model based on machine learning. Moreover, based on the characteristics of ant colony intelligent algorithm optimization, a method of using ant colony algorithm to solve the central node is proposed. In addition, this paper uses the ant colony algorithm to obtain the characteristic points in the study area and determine a reasonable number, and then combine the uniform grid to select some non-characteristic points as the central node of the core function, and finally use the central node with a reasonable distribution for modeling. Finally, this paper designs experiments to verify the performance of the model constructed in this paper and combines mathematical statistics to visually display the experimental results using tables and graphs. The research results show that the performance of the model constructed in this paper is good.

Neuronal Communication Genetic Algorithm-Based Inductive Learning

2021 ◽  
pp. 244-259
Author(s):  
Abdiya Alaoui ◽  
Zakaria Elberrichi
Keyword(s):  
Machine Learning ◽  
Genetic Algorithm ◽  
Predictive Accuracy ◽  
Inductive Learning ◽  
Learning Algorithms ◽  
Hybrid Approach ◽  
Machine Learning Algorithms ◽  
Rule Based ◽  
Neuronal Communication ◽  
Robust Model

The development of powerful learning strategies in the medical domain constitutes a real challenge. Machine learning algorithms are used to extract high-level knowledge from medical datasets. Rule-based machine learning algorithms are easily interpreted by humans. To build a robust rule-based algorithm, a new hybrid metaheuristic was proposed for the classification of medical datasets. The hybrid approach uses neural communication and genetic algorithm-based inductive learning to build a robust model for disease prediction. The resulting classification models are characterized by good predictive accuracy and relatively small size. The results on 16 well-known medical datasets from the UCI machine learning repository shows the efficiency of the proposed approach compared to other states-of-the-art approaches.

Benchmarking the Performance of a Machine Learning Classifier Enabled Multiobjective Genetic Algorithm on Six Standard Test Functions

2017 ◽  
Author(s):  
Kayla Zeliff ◽  
Walter Bennette ◽  
Scott Ferguson
Keyword(s):  
Machine Learning ◽  
Genetic Algorithm ◽  
Objective Function ◽  
Hybrid Approach ◽  
Pareto Frontier ◽  
Population Diversity ◽  
Benchmark Problems ◽  
Multiobjective Genetic Algorithm ◽  
Learning Classifier ◽  
Function Calls

Previous work tested a multi-objective genetic algorithm that was integrated with a machine learning classifier to reduce the number of objective function calls. Four machine learning classifiers and a baseline “No Classifier” option were evaluated. Using a machine learning classifier to create a hybrid multiobjective genetic algorithm reduced objective function calls by 75–85% depending on the classifier used. This work expands the analysis of algorithm performance by considering six standard benchmark problems from the literature. The problems are designed to test the ability of the algorithm to identify the Pareto frontier and maintain population diversity. Results indicate a tradeoff between the objectives of Pareto frontier identification and solution diversity. The “No Classifier” baseline multiobjective genetic algorithm produces the frontier with the closest proximity to the true frontier while a classifier option provides the greatest diversity when the number of generations is fixed. However, there is a significant reduction in computational expense as the number of objective function calls required is significantly reduced, highlighting the advantage of this hybrid approach.

scholarly journals A Novel Active Optimization Approach for Rapid and Efficient Design Space Exploration Using Ensemble Machine Learning

2019 ◽  
Author(s):  
Opeoluwa Owoyele ◽  
Pinaki Pal
Keyword(s):  
Machine Learning ◽  
Genetic Algorithm ◽  
Active Learning ◽  
Design Space ◽  
Particle Swarm ◽  
Global Optimum ◽  
Optimization Techniques ◽  
Machine Learning Algorithms ◽  
Ic Engine ◽  
Weak Learner

Abstract In this work, a novel design optimization technique based on active learning, which involves dynamic exploration and exploitation of the design space of interest using an ensemble of machine learning algorithms, is presented. In this approach, a hybrid methodology incorporating an explorative weak learner (regularized basis function model) which fits high-level information about the response surface, and an exploitative strong learner (based on committee machine) that fits finer details around promising regions identified by the weak learner, is employed. For each design iteration, an aristocratic approach is used to select a set of nominees, where points that meet a threshold merit value as predicted by the weak learner are selected to be evaluated using expensive function evaluation. In addition to these points, the global optimum as predicted by the strong learner is also evaluated to enable rapid convergence to the actual global optimum once the most promising region has been identified by the optimizer. This methodology is first tested by applying it to the optimization of a two-dimensional multi-modal surface. The performance of the new active learning approach is compared with traditional global optimization methods, namely micro-genetic algorithm (μGA) and particle swarm optimization (PSO). It is demonstrated that the new optimizer is able to reach the global optimum much faster, with a significantly fewer number of function evaluations. Subsequently, the new optimizer is also applied to a complex internal combustion (IC) engine combustion optimization case with nine control parameters related to fuel injection, initial thermodynamic conditions, and in-cylinder flow. It is again found that the new approach significantly lowers the number of function evaluations that are needed to reach the optimum design configuration (by up to 80%) when compared to particle swarm and genetic algorithm-based optimization techniques.

A Hybrid Approach for Feature Selection Based on Genetic Algorithm and Recursive Feature Elimination

2021 ◽  
Vol 12 (2) ◽  
pp. 17-38
Author(s):  
Pooja Rani ◽  
Rajneesh Kumar ◽  
Anurag Jain ◽  
Sunil Kumar Chawla
Keyword(s):  
Machine Learning ◽  
Genetic Algorithm ◽  
Feature Selection ◽  
Hybrid Approach ◽  
Feature Selection Method ◽  
Classification Systems ◽  
Recursive Feature Elimination ◽  
Support Vector ◽  
Sensitivity Specificity ◽  
Selection Of

Machine learning has become an integral part of our life in today's world. Machine learning when applied to real-world applications suffers from the problem of high dimensional data. Data can have unnecessary and redundant features. These unnecessary features affect the performance of classification systems used in prediction. Selection of important features is the first step in developing any decision support system. In this paper, the authors have proposed a hybrid feature selection method GARFE by integrating GA (genetic algorithm) and RFE (recursive feature elimination) algorithms. Efficiency of proposed method is analyzed using support vector machine classifier on the scale of accuracy, sensitivity, specificity, precision, F-measure, and execution time parameters. Proposed GARFE method is also compared to eight other feature selection methods. Results demonstrate that the proposed GARFE method has increased the performance of classification systems by removing irrelevant and redundant features.

Application of Genetic Algorithms to the Synthesis of Riser Configurations

2003 ◽  
Author(s):  
Luciano T. Vieira ◽  
Beatriz de S. L. P. de Lima ◽  
Alexandre G. Evsukoff ◽  
Breno P. Jacob
Keyword(s):  
Genetic Algorithm ◽  
Genetic Algorithms ◽  
Search Space ◽  
Optimization Techniques ◽  
Objective Functions ◽  
Random Choice ◽  
Large Space ◽  
Steel Catenary Riser ◽  
Deep Waters ◽  
Optimization Methodology

The purpose of this work is to describe the application of Genetic Algorithms in the search of the best configuration of catenary riser systems in deep waters. Particularly, an optimization methodology based on genetic algorithms is implemented on a computer program, in order to seek an optimum geometric configuration for a steel catenary riser in a lazy-wave configuration. This problem is characterized by a very large space of possible solutions; the use of traditional methods is an exhaustive work, since there is a large number of variables and parameters that define this type of system. Genetic algorithms are more robust than the more commonly used optimization techniques. They use random choice as a tool to guide a search toward regions of the search space with likely improvements. Some differences such as the coding of the parameter set, the search from a population of points, the use of objective functions and randomized operators are factors that contribute to the robustness of a genetic algorithm and result in advantages over traditional techniques. The implemented methodology has as baseline one or more criteria established by the experience of the offshore engineer. The implementation of an intelligent methodology oriented specifically to the optimization and synthesis of riser configurations will not only facilitate the work of manipulating a huge mass of data, but also assure the best alternative between all the possible ones, searching in a much larger space of possible solutions than classical methods.

Neuronal Communication Genetic Algorithm-Based Inductive Learning

2020 ◽  
Vol 13 (2) ◽  
pp. 141-154
Author(s):  
Abdiya Alaoui ◽  
Zakaria Elberrichi
Keyword(s):  
Machine Learning ◽  
Genetic Algorithm ◽  
Learning Strategies ◽  
Predictive Accuracy ◽  
Inductive Learning ◽  
Learning Algorithms ◽  
Hybrid Approach ◽  
Machine Learning Algorithms ◽  
Rule Based ◽  
High Level

The development of powerful learning strategies in the medical domain constitutes a real challenge. Machine learning algorithms are used to extract high-level knowledge from medical datasets. Rule-based machine learning algorithms are easily interpreted by humans. To build a robust rule-based algorithm, a new hybrid metaheuristic was proposed for the classification of medical datasets. The hybrid approach uses neural communication and genetic algorithm-based inductive learning to build a robust model for disease prediction. The resulting classification models are characterized by good predictive accuracy and relatively small size. The results on 16 well-known medical datasets from the UCI machine learning repository shows the efficiency of the proposed approach compared to other states-of-the-art approaches.

Hybrid Control System for Seismic Protection of the Benchmark Model

2012 ◽  
Vol 193-194 ◽  
pp. 801-804
Author(s):  
Jian Fang Fu ◽  
Xin Jian Feng ◽  
Jian Lin Zhang
Keyword(s):  
Genetic Algorithm ◽  
Control System ◽  
Control Systems ◽  
Hybrid Control ◽  
Seismic Protection ◽  
Control Parameters ◽  
Objective Functions ◽  
Benchmark Model ◽  
Damper Control ◽  
Hybrid Control System

In this paper, a 20-storey Benchmark model is used to compare the effects of various control systems, including Damper control system, TMD control system and Hybrid control system. In the course of the optimization, the displacement reducing coefficient (DRF) and the storey drift reducing coefficient (SDRF) are used as the objective functions to optimize the control parameters based on Genetic Algorithm (GA). Numerical results prove the advisability of the Hybrid control system of dampers and TMD, which can not only decrease the floor displacement, but also improve the mutation of the storey drift.

Economic Dispatch Using Genetic Algorithm Based Hybrid Approach

2006 ◽  
Author(s):  
Tahir Nadeem Malik ◽  
Aftab Ahmad ◽  
Shahab Khushnood
Keyword(s):  
Genetic Algorithm ◽  
Genetic Algorithms ◽  
Economic Dispatch ◽  
Hybrid Approach ◽  
Optimization Techniques ◽  
Characteristic Functions ◽  
Steam Turbines ◽  
Large Power ◽  
Hybrid Approaches ◽  
Power Economic

Power Economic Dispatch (ED) is vital and essential daily optimization procedure in the system operation. Present day large power generating units with multi-valves steam turbines exhibit a large variation in the input-output characteristic functions, thus non-convexity appears in the characteristic curves. Various mathematical and optimization techniques have been developed, applied to solve economic dispatch (ED) problem. Most of these are calculus-based optimization algorithms that are based on successive linearization and use the first and second order differentiations of objective function and its constraint equations as the search direction. They usually require heat input, power output characteristics of generators to be of monotonically increasing nature or of piecewise linearity. These simplifying assumptions result in an inaccurate dispatch. Genetic algorithms have used to solve the economic dispatch problem independently and in conjunction with other AI tools and mathematical programming approaches. Genetic algorithms have inherent ability to reach the global minimum region of search space in a short time, but then take longer time to converge the solution. GA based hybrid approaches get around this problem and produce encouraging results. This paper presents brief survey on hybrid approaches for economic dispatch, an architecture of extensible computational framework as common environment for conventional, genetic algorithm & hybrid approaches based solution for power economic dispatch, the implementation of three algorithms in the developed framework. The framework tested on standard test systems for its performance evaluation.

Sign in / Sign up

Export Citation Format

Share Document