The problem of sailing craft route optimisation involves considering a range of stochastic factors, from weather model used to prediction of performance and structural failure. This research introduces the sailing craft routing problem as a bi-level optimisation problem, with the leader problem being the route optimisation problem and the follower problem being the speed attainable at different reliability levels. The key issue is of how to model these stochastic factors. I use Julia exclusively due to its speed and the ability to rapidly develop code. In this talk I will introduce how I generate probabilistic temporal weather models. From these I am able to generate weather scenarios. To reduce computational running time I identify a representative collection of scenarios through comparing based on comparing probability metrics, solved using JuMP. I then show how I use monte carlo simulations to model the reliability of a sailing craft and then optimise the route using a recursive algorithm. This talk will demonstrate how I solve many optimisation and statistical problems which are common across engineering and scientific disciplines with respect to modelling and solving stochastic optimisation problems. As a result it has broad appeal across the scientific modelling community with a specific application in an unusual and interesting area.
