About this Project

What is it?

This project is a machine learning approach to predicting wind speeds in Ontario. It then takes these predicted wind speeds, and tries to optimize the placement of wind turbines within a specified area.

Why did I make this?

I have never worked with machine learning before, so I wanted to learn more about it by creating something with it.

I also wanted to challenge myself by solving a real world unique problem.

Who am I?

My name is Alexander, and I am a fourth year university student from Ontario! I am currently on my way to complete a double bachelors degree program of Computer Science and Business Administration.

Check out my Github and LinkedIn!

How did I make this?

Model

I used PyTorch to create my neural network model.

Data
Input Data

I created all of the input data myself. Each data point that was fed to the model included:

  • Latitude/Longitude
  • Elevation (above sea level)
  • Closest distance to a major body of water (i.e. The Great Lakes, Hudson Bay)
  • Closest distance to a major city in Ontario (and its population density)
  • If its within the Canadian Shield
  • Terrain Type (i.e. flat land, rocky, etc.)
Label Data

I luckily found a great website (Global Wind Atlas) that essentially would give you the Mean Wind Speed (m/s) for any coordinate on land/coast. Without this, I wouldn't be able to train my model, so check it out it's pretty cool!

Website

This website was created on the Django framework, with a simple combination of jQuery, and Bootstrap.