Upload folder using huggingface_hub
Browse files- README.md +30 -3
- ev_energy_model.joblib +3 -0
- requirements.txt +6 -0
README.md
CHANGED
|
@@ -1,3 +1,30 @@
|
|
| 1 |
-
|
| 2 |
-
|
| 3 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
# ⚡ EV Energy Consumption Predictor (ANN + Physical Filtering)
|
| 2 |
+
|
| 3 |
+
Este modelo predice el **consumo energético de un Vehículo Eléctrico (EV)** en kWh dado un conjunto de variables físicas y operacionales.
|
| 4 |
+
Se basa en un pipeline completo de:
|
| 5 |
+
|
| 6 |
+
- limpieza física del dataset (SoC_diff > 0, charging duration > 0, efficiency bounds)
|
| 7 |
+
- feature engineering derivado de parámetros de carga
|
| 8 |
+
- selección de features según correlación
|
| 9 |
+
- escalado (StandardScaler)
|
| 10 |
+
- red neuronal artificial (ANN) entrenada para regresión
|
| 11 |
+
|
| 12 |
+
---
|
| 13 |
+
|
| 14 |
+
## 📘 Uso del Modelo
|
| 15 |
+
|
| 16 |
+
### Entrada esperada (JSON)
|
| 17 |
+
Debe incluir las mismas columnas utilizadas para el entrenamiento:
|
| 18 |
+
|
| 19 |
+
```json
|
| 20 |
+
{
|
| 21 |
+
"Battery Capacity (kWh)": 75,
|
| 22 |
+
"SoC_diff": 40,
|
| 23 |
+
"Charging Duration (hours)": 1.5,
|
| 24 |
+
"Energy_est_SoC": 30.0,
|
| 25 |
+
"Charging_Rate": 20.0,
|
| 26 |
+
"Power_proxy": 25.0,
|
| 27 |
+
"Charge_Efficiency": 0.92,
|
| 28 |
+
"Energy_per_SoC": 0.75,
|
| 29 |
+
"Vehicle Age (years)": 3
|
| 30 |
+
}
|
ev_energy_model.joblib
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:49a3c4944e3c2f495e020243e325bbbbb178fefb804507be88cfe6f37d5f516a
|
| 3 |
+
size 174463
|
requirements.txt
ADDED
|
@@ -0,0 +1,6 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
pandas
|
| 2 |
+
numpy
|
| 3 |
+
scikit-learn
|
| 4 |
+
tensorflow
|
| 5 |
+
joblib
|
| 6 |
+
matplotlib
|