Le 14/11/2023 par idoumghar :

Internship: Combining Artificial Intelligence Approaches and Metaheuristics to Solve Zeolite Structure Determination Problem

Introduction

Metaheuristics are a class of optimization algorithms designed to tackle complex problems by providing efficient and adaptable search strategies. Their primary aim is to find near-optimal solutions in a reasonable amount of time, making them invaluable in various domains, including engineering, logistics, finance, and science, where they can address real-world challenges that often involve high computational complexity. One of these computationally complex problems is the zeolite structure determination problem. Zeolites are crystalline microporous materials with a wide range of applications, including catalysis, adsorption, and ion exchange. Understanding and characterizing zeolite structures is crucial for optimizing their performance in various industrial processes. One of the challenges in zeolite research is the determination of their crystal structures, which involves solving complex mathematical optimization problems. This internship proposal aims to design hybrid AI-based metaheuristics and explore their application to efficiently solve zeolite structures, offering innovative solutions to a long-standing problem in materials science.

Objectives

The primary objectives of this master internship proposal are as follows:

1. Study the AI-based hybrid metaheuristics.
2. Investigate and implement metaheuristic algorithms for the purpose of solving zeolite structures.
3. Compare the performance of various metaheuristic techniques in terms of efficiency and accuracy for zeolite structure determination.
4. Validate the results obtained from the developed algorithms against experimental data to assess their reliability and applicability.

Methodology

The internship will be organized into the following stages:

Stage 1: Literature Review (1 months)

1. Develop a comprehensive understanding of metaheuristic algorithms and their potential application.
2. Review the existing literature on zeolite structures, crystallography, and structure analysis techniques used in materials science.
3. Identify the gaps and challenges in the current methods for solving zeolite structures.

Stage 2: Algorithm Implementation (3 months)

1. Select and implement various metaheuristic algorithms for solving zeolite structures.
2. Develop or adapt optimization functions to represent zeolite structure determination as an optimization problem.
3. Propose new AI-based hybrid metaheuristics to improve their performance.

Stage 3: Comparative Analysis (2 months)

1. Apply the implemented algorithms to a dataset of known zeolite structures and evaluate their performance in terms of convergence, accuracy, and computational efficiency.
2. Compare the results with existing methods used in zeolite structure determination.

Expected Outcomes

1. A set of implemented metaheuristic algorithms tailored for solving zeolite structures.
2. Comparative analysis of algorithm performance in terms of efficiency and accuracy.
3. Validation of algorithm results through experimental data, demonstrating their practicality in materials science.

Timeline

This internship is expected to span 6 months, beginning in 02/2024 and ending in 08/2024.

Mentorship and Resources

The internship will take place within the collaboration between IRIMAS and IS2M, which are two laboratories of Université de Haute Alsace, Mulhouse, France. Access to laboratory facilities, computational resources, and necessary software tools will be provided throughout the internship.

Profile

This internship proposal offers an exciting opportunity for a motivated and research-oriented master's student with strong skills in C++ and Python alongside a basic knowledge of optimization and artificial intelligence techniques. The intern is expected to make contributions to the field of materials science by combining the power of AI approaches and metaheuristics to solve the zeolite structure determination problem.

Contact

Mahmoud GOLABI, mahmoud.golabi@uha.fr
Taylan ORS, taylan.ors@uha.fr 
Mokhtar ESSAID mokhtar.essaid@uha.fr
Jean-Louis PAILLAUD,  jean-louis.paillaud@uha.fr
Lhassane IDOUMGHAR, lhassane.idoumghar@uha.fr 

Address

12, Rue des Frères Lumières, 68093, Mulhouse, FRANCE.