by Matheus Bernardelli de Moraes, Guilherme Palermo Coelho, presented at 2022 IEEE Congress on Evolutionary Computation (CEC), September 2022.
Abstract—Many real-world optimization problems involve time-consuming fitness evaluation. To reduce the computational cost of expensive evaluations, researchers have been developing surrogate models to approximate the objective function values of unevaluated candidate solutions. However, most of the research has been developed for continuous optimization problems, while only a few of them address surrogate modeling for expensive multi-objective Combinatorial Optimization Problems (COPs). COPs have inherently different challenges than continuous opti- mization. For example, (i) many COPs have categorical and nom- inal decision variables; (ii) they often require the combination of both global and local search mechanisms; and (iii) some of them have constraints that make them NP-hard problems, which makes them even more difficult to solve with a reasonable number of fitness evaluations. To address these issues, this paper proposes a surrogate-assisted evolutionary algorithm that combines the decomposition-based algorithm MOEA/D, Tabu Local Search, and Random Forest as a surrogate model to approximate the objective function of unevaluated individuals on multi-objective COPs. Experiments were conducted on constrained and uncon- strained well-known multi-objective combinatorial optimization benchmark problems. The experimental results demonstrate that the proposed design outperforms state-of the-art algorithms without violating the restrictions in the number of objective function evaluations, which indicates that it may be suitable for real-world expensive multi-objective COPs.
by Matheus Bernardelli de Moraes, Guilherme Palermo Coelho, published at Expert Systems with Applications, September 2022, Vol. 202, 117251.
Expensive multi-objective combinatorial optimization problems have constraints in the number of objective function evaluations due to time, financial, or resource restrictions. As most combinatorial problems, they are subject to a high number of duplicated solutions. Given the fact that expensive environments limit the number of objective function evaluations, the existence of duplicated solutions heavily impacts the optimization process due to poor diversity and low convergence speed. This paper proposes the Novel-First Tabu Search, a greedy-strategy mechanism that uses Knowledge-Assisted Local Search methods to preserve the population diversity and increase the exploration and exploitation ability of MOEA/D. Experiments are conducted on constrained, unconstrained, multimodal, deceptive, linear, convex, and non-convex Pareto Front multi-objective combinatorial optimization benchmark problems. This paper also conducts an experiment on the real-world, expensive problem of Well Placement Optimization using a benchmark case based on the Namorado oil field, located in the Campos Basin, Brazil. The experimental results and performance comparison with state-of-the-art algorithms demonstrate that the proposed design significantly preserves diversity and increases convergence without violating the constraint in the number of objective function evaluations.
Abouzar Mirzaei-Paiaman, Susana M.G. Santos, Denis J. Schiozer, published at Journal of Petroleum Science and Engineering, November 2022, Vol. 218, 111005.
Determination of optimal quantity and location of wells is a crucial step in any field development project. Multiple challenges exist in the real-world projects, such as gigantic multidimensional search space and complex parameterization of the problem. To address these challenges, we present an iterative sequential optimization framework that consists of multiple stages for optimizing number and location of producers and injectors. Once the initial guesses have been established, the first optimization stage begins in order to determine the optimal number of producers and injectors. Next, a two-stage process is used to determine the best location for each well. By optimizing the number and location of wells, well type is optimized implicitly and does not need to be included in the optimization process as an individual decision variable. To consider the dependency and coupling between parameters, the entire optimization process is repeated iteratively up to a stopping point. The proposed approach is highly flexible and can be extended to include other potential decision variables of field development projects. We test the proposed methodology on UNISIM–II–D, a realistic benchmark case that represents pre-salt reservoirs in deep offshore Brazil, which are light-oil fractured carbonates with large volumes of CO2-rich associated gas. In our application, all producers and injectors are equipped with internal control valves (ICVs), all produced gas is reinjected into the reservoir, and water-alternating-gas with CO2 (WAG-CO2) is used to increase oil recovery and ensure that the produced CO2, as a harmful greenhouse gas, can be safely discarded under- ground. We use the iterative discrete Latin hypercube (IDLHC) optimization algorithm. To achieve a robust decision, we consider subsurface, operational and economic uncertainties by considering nine representative reservoir models and three economic scenarios. Based on the results, the proposed framework can provide an appropriate solution containing optimal quantity of wells, their locations and types in realistic cases with large search spaces and complex parameterizations.
by Paulo Henrique Ranazzi, Xiaodong Luo, Marcio Augusto Sampaio, published at Journal of Petroleum Science and Engineering, August 2022, Vol. 215 (Part A), 110589.
In present days, Iterative ensemble smoothers (IES) are among the main methods to perform ensemble-based history matching in petroleum reservoirs. Generally, some localization technique is applied to the IES to pre- vent ensemble collapse, which is the consequence of an excessive reduction of the posterior ensemble variance. When the standard distance-based localization is applied, the assimilation of non-local parameters is difficult, and besides that, this kind of methodology has also several intrinsic parameters that need to be defined before the assimilation. In contrast, adaptive localization methods aim to overcome the noticed problems of distance-based localization, by using some statistical method to define the localization. This article proposes a novel adaptive localization scheme, on top of two preexisting techniques: pseudo-optimal and correlation-based localizations. The motivation here is to further improve the adaptive localization scheme, by combining the strengths of these two preexisting techniques. The efficacy of the proposed localization scheme is tested in one 2D and one 3D case studies, whereas the latter case study involves a field-scale reservoir model with both local and non-local pa- rameters, which often impose challenges on the conventional localization schemes. In comparison to other evaluated localization schemes, our results indicate that the proposed adaptive localization scheme achieves improved history matching performance.
Atualização (21/10/2022): diante da grande procura e como o número de vagas era limitado, as inscrições para a 4th EPIC Conference estão encerradas.
O EPIC gostaria de convidar estudantes de pós-graduação, pesquisadores e profissionais da área de engenharia de reservatórios para a 4th. EPIC Conference, evento anual voltado à divulgação dos trabalhos em desenvolvimento no centro. A 4th. EPIC Conference ocorrerá entre os dias 07 e 09 de novembro de 2022* no campus da Universidade Estadual de Campinas (UNICAMP), localizado em Campinas – SP – Brasil.
As atividades da 4th EPIC Conference estarão distribuídas em dois locais:
* O minicurso intitulado “Electric Submersible Pumps (ESPs): Fundamentals, Failure Analysis & Reliability, and Emulsion as a Flow Assurance Issue” terá uma sessão extra no dia 10 de novembro de 2022, às 8:30.