File size: 5,123 Bytes
991ba9c
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
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
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
# Energy ML API Documentation

## Overview
This API provides machine learning predictions for industrial energy management with three separate endpoints for different models.

## Base URL
```
https://your-space-name.hf.space
```

## Endpoints

### 1. Energy Prediction - Random Forest
**Endpoint:** `/api/predict` (Tab 0)  
**Method:** POST  
**Description:** Predicts daily average energy consumption using Random Forest

#### Request Body
```json
[
  {
    "data": "2025-01-01",
    "boosting": 0.0,
    "cor": "incolor",
    "espessura": 8.0,
    "extracao_forno": 750.0,
    "porcentagem_caco": 15.0,
    "extracao_boosting": 1.5
  }
]
```

#### Response
```json
[
  {
    "data": "01-01-2025",
    "predicted_energy": 5.234,
    "note": "Add boosting energy separately to get final consumption"
  }
]
```

### 2. Energy Prediction - XGBoost
**Endpoint:** `/api/predict` (Tab 1)  
**Method:** POST  
**Description:** Predicts daily average energy consumption using XGBoost (best model)

#### Request Body
```json
[
  {
    "data": "2025-01-01",
    "boosting": 0.0,
    "cor": "incolor",
    "espessura": 8.0,
    "extracao_forno": 750.0,
    "porcentagem_caco": 15.0,
    "extracao_boosting": 1.5
  }
]
```

#### Response
```json
[
  {
    "data": "01-01-2025",
    "predicted_energy": 5.187,
    "note": "Add boosting energy separately to get final consumption"
  }
]
```

### 3. Threshold Detection
**Endpoint:** `/api/predict` (Tab 2)  
**Method:** POST  
**Description:** Predicts probability of exceeding 8300 and 9000 consumption thresholds

#### Request Body
```json
{
  "data": "2025-01-01",
  "cor": "incolor",
  "espessura": 8.0,
  "ext_boosting": 1.5,
  "extracao_forno": 750.0,
  "porcentagem_caco": 15.0,
  "prod_e": 1,
  "prod_l": 0,
  "autoclave": 1
}
```

#### Response
```json
{
  "predictions": {
    "prediction_1": [
      {
        "datetime": "2025-01-01",
        "threshold": 8300,
        "probabilidade_de_estouro": 0.1234,
        "estouro_previsto": 0
      }
    ],
    "prediction_2": [
      {
        "datetime": "2025-01-01",
        "threshold": 9000,
        "probabilidade_de_estouro": 0.0567,
        "estouro_previsto": 0
      }
    ]
  }
}
```

## Input Parameters

### Energy Prediction Models
| Parameter | Type | Required | Description |
|-----------|------|----------|-------------|
| `data` | string | Yes | Date in YYYY-MM-DD format |
| `boosting` | float | Yes | Always set to 0.0 (compatibility) |
| `cor` | string | Yes | Glass color: "incolor", "verde", "cinza", "bronze" |
| `espessura` | float | Yes | Glass thickness (mm) |
| `extracao_forno` | float | Yes | Furnace extraction rate |
| `porcentagem_caco` | float | Yes | Glass cullet percentage (%) |
| `extracao_boosting` | float | Yes | External boosting parameter |

### Threshold Detection Model
| Parameter | Type | Required | Description |
|-----------|------|----------|-------------|
| `data` | string | Yes | Date in YYYY-MM-DD format |
| `cor` | string | Yes | Glass color: "incolor", "verde", "cinza", "bronze" |
| `espessura` | float | Yes | Glass thickness (mm) |
| `ext_boosting` | float | Yes | External boosting parameter |
| `extracao_forno` | float | Yes | Furnace extraction rate |
| `porcentagem_caco` | float | Yes | Glass cullet percentage (%) |
| `prod_e` | int | Yes | Production E (0 or 1) |
| `prod_l` | int | Yes | Production L (0 or 1) |
| `autoclave` | int | Yes | Autoclave usage (0 or 1) |

## Usage Examples

### Python
```python
import requests
import json

# Energy Prediction (XGBoost)
url = "https://your-space-name.hf.space/api/predict"
headers = {"Content-Type": "application/json"}

data = [
    {
        "data": "2025-01-01",
        "boosting": 0.0,
        "cor": "verde",
        "espessura": 10.0,
        "extracao_forno": 849.0,
        "porcentagem_caco": 45.0,
        "extracao_boosting": 54.0
    }
]

response = requests.post(url, json=data, headers=headers)
prediction = response.json()
print(f"Predicted energy: {prediction[0]['predicted_energy']} MWh")
```

### cURL
```bash
# Threshold Detection
curl -X POST https://your-space-name.hf.space/api/predict \
  -H "Content-Type: application/json" \
  -d '{
    "data": "2025-01-01",
    "cor": "incolor",
    "espessura": 8.0,
    "ext_boosting": 1.5,
    "extracao_forno": 750.0,
    "porcentagem_caco": 15.0,
    "prod_e": 1,
    "prod_l": 0,
    "autoclave": 1
  }'
```

## Model Performance
- **Random Forest Energy**: R² = 0.50, MAE = 0.24 MWh
- **XGBoost Energy**: R² = 0.53, MAE = 0.24 MWh (recommended)
- **Threshold Detection**: 98.9% accuracy (8300), 100% accuracy (9000)

## Notes
- **Energy models predict base consumption** - add actual boosting energy separately
- **No temporal features** - models work for any date/time period
- **Real-time prediction** - sub-second response times
- **Batch processing** - energy models accept arrays of inputs
- **Threshold models** output probabilities (0-1) and binary predictions (>0.5)

## Error Handling
- Invalid JSON: Returns error message
- Missing parameters: Uses default values where possible
- Model errors: Returns error description with context