blob: e23bc4ad85b32345b7382014e10b4adca4617b97 (
plain)
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
|
# Day-Ahead Wind Speed Forecasting for Nord Pool Bidding
Pluto notebook accompanying the individual weather data paper for *Introduction to Business Analytics* (EBC1042), Maastricht University School of Business and Economics.
The paper proposes a day-ahead wind speed forecasting tool for the Alajärvi Wind Farm Complex (Ilmatar Energy), using daily FMI station data from Meteostat.
## Models
| Model | Description |
|-------|-------------|
| M0 | Persistence baseline (tomorrow = today) |
| M1 | Static OLS on Möksy plus the four inner-ring stations (10 lags) |
| M2 | Rolling 365-day OLS on Möksy plus inner and outer rings (18 lags) |
All models predict day *t* from observations at *t − 1* and *t − 2*. Calibration runs January 2021 to October 2025, and the holdout runs October 2025 to October 2026.
## Running the notebook
Requires [Julia](https://julialang.org/downloads/) and [Pluto.jl](https://plutojl.org/).
```bash
julia -e 'using Pluto; Pluto.run(notebook="Paper.jl")'
```
Pluto installs the required packages (GLM, DataFrames, LinearAlgebra) automatically on first run.
## Data
Daily wind speed observations from FMI synoptic stations, retrieved via [Meteostat](https://meteostat.net/). Place the data files in the project folder as described in the notebook.
## Files
- `Paper.jl`: the Pluto notebook with the full analysis
- `README.md`: this file
- `LICENSE`: license terms
## Author
Mikka Anton Petterson
|
