How do you design the most cost efficient jacket (steel sub-structure supporting a topside)?
To find the optimal jacket design entails a minimization of the steel weight, while ensuring the structural integrity of the jacket as well as ensuring that it can be efficiently constructed.
Kvaerner has made use of a genetic algorithm to optimise the jacket geometry, but while the algorithm itself is efficient, the third-party software used to evaluate the structures is more time consuming. This results in an interest of reducing the number of generations needed.
Theoretical work includes:
1. Literature study of relevant methods used in structural design
2. Literature study of bio-inspired AI, and more generally stochastic optimization techniques, used in automated engineering
3. Literature study on methods used to avoid early convergence and how to handle costly fitness evaluation
The starting point will be optimization via existing GA used by Kværner today. Calculating fitness is the main bottleneck today, so reducing the number of generations, reducing the population size, or other improved ways to handle the cost of fitness evaluation will be beneficial. To do so, the student group may, for example, study, implement, and test improvements in evolutionary algorithms, such as in crossover and mutation. Supporting mechanism will also be looked into. The group may also look into other options, such as avoiding the calculation of fitness every generation using the 3rd party program in use today.
In this project, the joint interests of sponsor, advisor and student(s) come together. Typically, the project will be based on the problem described by the sponsor AND previous research by Prof. Ole Jakob Mengshoel AND interests of students. If only one or two of these are present, there is no basis for a project.
Please send email(s) to sponsor and/or advisor if you're interested in this project.
Ole Jakob Mengshoel is (more or less) following Keith Downing's selection process for master students - read this: