hareshchander commited on
Commit
70171eb
·
verified ·
1 Parent(s): d7a5947

Update app.py

Browse files
Files changed (1) hide show
  1. app.py +58 -13
app.py CHANGED
@@ -1,12 +1,41 @@
1
  import streamlit as st
2
  import pandas as pd
3
  import plotly.express as px
 
 
 
4
  from app_backend import fetch_weather, generate_synthetic_data, optimize_load
5
 
6
  # Constants
7
  API_KEY = "84e26811a314599e940f343b4d5894a7"
8
  DEFAULT_LOCATION = "New York"
9
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
10
  # Sidebar
11
  st.sidebar.title("Smart Grid Dashboard")
12
  location = st.sidebar.text_input("Enter Location", DEFAULT_LOCATION)
@@ -21,45 +50,61 @@ if weather:
21
  # Main dashboard title
22
  st.title("Real-Time Smart Grid Dashboard")
23
 
24
- # **User upload section**
25
  uploaded_file = st.file_uploader("Upload your Smart Grid CSV file", type=["csv"])
26
 
27
  if uploaded_file is not None:
28
- # Read uploaded CSV into DataFrame
29
  data = pd.read_csv(uploaded_file)
30
  st.success("File uploaded successfully!")
31
  else:
32
- # No file uploaded, fallback to synthetic data
33
  st.info("No file uploaded. Using synthetic data for demonstration.")
34
  data = generate_synthetic_data()
35
 
36
- # Check if the data contains required columns
37
- required_columns = ["timestamp", "load_demand_kwh", "solar_output_kw", "wind_output_kw", "grid_health"]
38
  if all(col in data.columns for col in required_columns):
39
 
40
- # Plot load demand
41
- fig_load = px.line(data, x="timestamp", y="load_demand_kwh", title="Load Demand Over Time")
42
  st.plotly_chart(fig_load)
43
 
44
- # Renewable energy contribution plot
45
  fig_renew = px.bar(
46
  data,
47
  x="timestamp",
48
- y=["solar_output_kw", "wind_output_kw"],
49
  title="Renewable Energy Contributions",
50
  labels={"value": "Power (kW)", "variable": "Energy Source"}
51
  )
52
  st.plotly_chart(fig_renew)
53
 
54
- # Grid health overview
55
  st.subheader("Grid Health Overview")
56
  grid_health_counts = data["grid_health"].value_counts()
57
  st.bar_chart(grid_health_counts)
58
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
59
  # Optimization recommendations based on last timestamp
60
- current_demand = data["load_demand_kwh"].iloc[-1]
61
- current_solar = data["solar_output_kw"].iloc[-1]
62
- current_wind = data["wind_output_kw"].iloc[-1]
63
  recommendation = optimize_load(current_demand, current_solar, current_wind)
64
 
65
  st.subheader("Recommendations")
 
1
  import streamlit as st
2
  import pandas as pd
3
  import plotly.express as px
4
+ import torch
5
+ import torch.nn as nn
6
+ import joblib
7
  from app_backend import fetch_weather, generate_synthetic_data, optimize_load
8
 
9
  # Constants
10
  API_KEY = "84e26811a314599e940f343b4d5894a7"
11
  DEFAULT_LOCATION = "New York"
12
 
13
+ # Define your PyTorch model architecture (match what you trained)
14
+ class GridModel(nn.Module):
15
+ def __init__(self):
16
+ super(GridModel, self).__init__()
17
+ self.layers = nn.Sequential(
18
+ nn.Linear(5, 16),
19
+ nn.ReLU(),
20
+ nn.Linear(16, 8),
21
+ nn.ReLU(),
22
+ nn.Linear(8, 1),
23
+ nn.Sigmoid()
24
+ )
25
+ def forward(self, x):
26
+ return self.layers(x)
27
+
28
+ # Load scaler and model once
29
+ @st.cache_resource
30
+ def load_model_and_scaler():
31
+ scaler = joblib.load("scaler.pkl")
32
+ model = GridModel()
33
+ model.load_state_dict(torch.load("model.pth", map_location=torch.device('cpu')))
34
+ model.eval()
35
+ return model, scaler
36
+
37
+ model, scaler = load_model_and_scaler()
38
+
39
  # Sidebar
40
  st.sidebar.title("Smart Grid Dashboard")
41
  location = st.sidebar.text_input("Enter Location", DEFAULT_LOCATION)
 
50
  # Main dashboard title
51
  st.title("Real-Time Smart Grid Dashboard")
52
 
53
+ # User upload section
54
  uploaded_file = st.file_uploader("Upload your Smart Grid CSV file", type=["csv"])
55
 
56
  if uploaded_file is not None:
 
57
  data = pd.read_csv(uploaded_file)
58
  st.success("File uploaded successfully!")
59
  else:
 
60
  st.info("No file uploaded. Using synthetic data for demonstration.")
61
  data = generate_synthetic_data()
62
 
63
+ required_columns = ["timestamp", "voltage", "frequency", "load", "solar_gen", "wind_gen", "grid_health"]
 
64
  if all(col in data.columns for col in required_columns):
65
 
66
+ # Show plots (adjust columns if needed)
67
+ fig_load = px.line(data, x="timestamp", y="load", title="Load Over Time")
68
  st.plotly_chart(fig_load)
69
 
 
70
  fig_renew = px.bar(
71
  data,
72
  x="timestamp",
73
+ y=["solar_gen", "wind_gen"],
74
  title="Renewable Energy Contributions",
75
  labels={"value": "Power (kW)", "variable": "Energy Source"}
76
  )
77
  st.plotly_chart(fig_renew)
78
 
 
79
  st.subheader("Grid Health Overview")
80
  grid_health_counts = data["grid_health"].value_counts()
81
  st.bar_chart(grid_health_counts)
82
 
83
+ # Prepare features for model prediction
84
+ features = data[['voltage', 'frequency', 'load', 'solar_gen', 'wind_gen']].values
85
+ X_scaled = scaler.transform(features)
86
+ inputs = torch.tensor(X_scaled, dtype=torch.float32)
87
+
88
+ # Run model prediction
89
+ with torch.no_grad():
90
+ outputs = model(inputs).numpy().flatten()
91
+
92
+ # Compute average output as stability metric
93
+ avg_output = outputs.mean()
94
+ if avg_output < 0.4:
95
+ status = "Unstable"
96
+ elif avg_output < 0.7:
97
+ status = "Warning"
98
+ else:
99
+ status = "Stable"
100
+
101
+ st.subheader("Grid Stability Prediction")
102
+ st.write(f"Predicted Grid Stability Status: **{status}** (Confidence: {avg_output:.2f})")
103
+
104
  # Optimization recommendations based on last timestamp
105
+ current_demand = data["load"].iloc[-1]
106
+ current_solar = data["solar_gen"].iloc[-1]
107
+ current_wind = data["wind_gen"].iloc[-1]
108
  recommendation = optimize_load(current_demand, current_solar, current_wind)
109
 
110
  st.subheader("Recommendations")