Moving from traditional energy plans powered by fossils fuels to unlimited renewable energy subscriptions allows for instant access to clean energy without heavy investment in infrastructure like solar panels, for example.
One clean energy source that has been gaining popularity around the world is wind turbines. Turbines are massive structures that are strategically placed in perpetually windy places to generate the most energy. Wind energy is generated when the power of the atmosphere’s airflow is harnessed to create electricity. Wind turbines do this by capturing the kinetic energy of the wind. Factors such as temperature, wind direction, turbine status, weather, blade length, etc. influence the amount of power generated.
Task
Predict the power that is generated (in KW/h) based on the various features provided in the dataset.
Data description
The dataset folder contains the following files:
- train.csv: 28200 x 22
- test.csv: 12086 x 21
- sample_submission.csv: 5 x 3
The dataset contains the following columns:
| Column name | Description |
| tracking_id | Represents a unique identification number of a windmill |
| datetime | Represents the date and time of a record |
| wind_speed(m/s) | Represents the speed of wind (in meter per second) |
| atmospheric_temperature(°C) | Represents the temperature (in degree Celcius) of a town or village that the windmill is present in |
| shaft_temperature(°C) | Represents the temperature of the shaft (in degree Celcius) |
| blades_angle(°) | Represents the angle of the blades of a wind turbine (in degrees) |
| gearbox_temperature(°C) | Represents the temperature of a gearbox (in degree Celcius) |
| engine_temperature(°C) | Represents the temperature of an engine (in degree Celcius) |
| motor_torque(N-m) | Represents the torque of a motor (in Newton meter) |
| generator_temperature(°C) | Represents the temperature of a generator (in degree Celcius) |
| atmospheric_pressure(Pascal) | Represents the atmospheric pressure (in Pascals) in that area |
| area_temperature(°C) | Represents the temperature (in degree Celcius) of the area within a 100 m radius of the windmill |
| windmill_body_temperature(°C) | Represents the temperature of the body of a windmill (in degree Celcius) |
| wind_direction(°) | Represents the direction of the wind (in degrees) |
| resistance(ohm) | Represents the resistance against the wind |
| rotor_torque(N-m) | Represents the torque of a rotor (in Newton meter) |
| turbine_status | Represents the status of the turbine (masked) |
| cloud_level |
Represents the following levels of the cloud in the sky on a particular day:
|
| blade_length(m) | Represents the length of the blades of a windmill (in meter) |
| blade_breadth(m) | Represents the breadth of the blades of a windmill (in meter) |
| windmill_height(m) | Represents the height of the blades of a windmill (in meter) |
| windmill_generated_power(kW/h) | Represents the power generated (in Kilo Watt per hour) |
Evaluation metric
score = max(0 ,100*metrics.r2_score(actual, predicted))Result submission guidelines
- The indexes are tracking_id and datetime.
- The target is the windmill_generated_power(kW/h) column.
- The submission file must be submitted in .csv format only.
- The size of this submission file must be 12086 x 3.
Note: Ensure that your submission file contains the following:
- Correct index values as per the test file
- Correct names of columns as provided in the sample_submission.csv file
