HOW FAST CAN A SAILBOAT GO: A CONSTRAINED OPTIMIZATION APPROACH

Physics Behind a Sailboat In some areas Hong Kong, to name a few, Repulse Bay, Sai Kung, etc.

η«‹ε€ε€œδΉ‹ζ’¦

A Prelude What is the symbolization of our dreams? What are the representation and abstraction of the characteristics and scenes appeared in our dreams?

η«‹ε€ε€œδΉ‹ζ’¦

A Prelude What is the symbolization of our dreams? What are the representation and abstraction of the characteristics and scenes appeared in our dreams?

GRADIENT DESCENT: NEWTON AND QUASI-NEWTON METHOD

Newton’s Method Different from gradient-descent (GD) method which tries to directly converge the design variable to a minimiser, Newton’s method uses an important characteristic of an extreme value – gradient of the function ($\nabla \mathbf{f}$) equals zero.

2D RANS SIMULATION OF FLOATING VAWT IN ROLL MOTION

Background A simple 2D Reynolds-Averaged Navier-Stokes (RANS) simulation of a floating vertical axis wind turbine (VAWT) in roll motion, sliced by a top view.

GRADIENT DESCENT: LINE SEARCH

Introduction of Gradient Descent Gradient Descent (GD) is an optimisation algorithm that does not need my introduction.