# Sailing Simulation ## Summary The application allows the agent to control a sailboat by turning left or right to navigate obstacles. ^d3afa9 Distance moved after every action is calculated based on the direction the boat is facing against the wind. ## Data Source The data is adapted from a Sailing Simulation by https://github.com/PPierzc/ai-learns-to-sail. The sailboat distance is calculated based on the direction against the wind. The reward for reaching the objective is +1 and conversely -1 for failing. ## elsciRL App Options ### Local Config The following table summarises the adapters available in the current application. | Config Name | Config Code | Description | |:------------:|:------------:|:----------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------:| | Easy River | easy | The simplest setting; fixed wind direction and a 25m distance to the goal line without hitting the walls. Observation precise to 2 decimal places, reward = 0.5 for sub-goal, -0.1 if the action limit of 100 is reached. | | | | | ### Adapters The following table summarises the adapters available in the current application. | Adapter Name | Adapter Code | Description | Output type | |:---------------:|:------------:|:-------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------:|:-----------:| | Default Numeric | default | States "x_angle" that denotes the horizontal position and angle against the wind are indexed based on the list of all possible states and then transformed to a tensor. | $tensor$ | | Simple Language | language | States "x_angle" are transformed to a description based on: 1) horizontal position, 2) side of river (beach or harbor side), 3) angle against the wind and 4) direction of previous action taken. | $tensor$ | ## Citation ```bibtex @phdthesis{Osborne2024, title = {Improving Real-World Reinforcement Learning by Self Completing Human Instructions on Rule Defined Language}, author = {Philip Osborne}, year = 2024, month = {August}, address = {Manchester, UK}, note = {Available at \url{https://research.manchester.ac.uk/en/studentTheses/improving-real-world-reinforcement-learning-by-self-completing-hu}}, school = {The University of Manchester}, type = {PhD thesis} } ```