(337bb) Computational Tools for Process Design

Over the course of my academic years, I have been focusing on the development and application of computational Process Systems Engineering (PSE) and machine learning methods to tackle various chemical engineering problems.

Research Interests:

My main research area of interest is in the development and application of computational tools for use within the chemical engineering process design/control/optimisation. During my MSc research project, I worked on applying machine learning tools for simultaneous design and control of chemical processes. The methodology leverages the ability to embed reinforcement learning (RL) controllers in a mixed-integer optimization problem for process design. Two case studies were taken from literature where the authors have implemented a multi-parametric approach towards this problem. This first considers the design of a mixing tank, while the second is of a heated CSTR. Our RL method was able to arrive at the same solution with the added benefit of controllers that can address disturbances better. The results of my MSc research project have been published and presented in several conferences.

Afterwards, I started my PhD on developing a framework for the identification of design space and quantification of process flexibility. I developed a Python package which enables the identification of design spaces based on the formation of geometrical hulls (alpha shape) and solving the flexibility analysis problem in proxy by exploiting the identified design space circumventing the need to solve complex optimization problems. This framework has been applied to protein A affinity chromatography for monoclonal antibody (mAb) capture. I quantified trade-offs between performance and flexibility which are key questions to be answered for process design. I have also looked into screening of protein A affinity resins in the context of flexibility-performance trade-offs with experimental validation from our collaborators. The framework has also been applied to post-combustion carbon capture pressure/vacuum-swing adsorption process (PVSA). We quantified performance-flexibility trade-offs which enhances process understanding and accelerates process development. Currently, I am working towards finishing the last chapter of my PhD which is on multi-parametric design space. The aim for this last chapter is to tackle higher dimensional problems by formulating the design space as a multi-parametric problem and using the multi-parametric solution as a proxy to solve the flexibility test problem.

I am interested in a career where I can apply my technical computational skills, solving wide variety of complex real-world problems, and learning new knowledge and skills to further enhance my personal and career development.