import streamlit as st import pandas as pd import plotly.express as px import torch import torch.nn as nn import numpy as np from app_backend import fetch_weather, generate_synthetic_data, optimize_load # ---------------------------- # Dummy PyTorch Model Definition # ---------------------------- class GridStabilityModel(nn.Module): def __init__(self): super(GridStabilityModel, self).__init__() self.net = nn.Sequential( nn.Linear(3, 16), nn.ReLU(), nn.Linear(16, 8), nn.ReLU(), nn.Linear(8, 1), nn.Sigmoid() ) def forward(self, x): return self.net(x) # Load or instantiate model model = GridStabilityModel() model.eval() # Set model to eval mode # Function to predict stability score using the model def predict_stability(data): # Select features for prediction features = data[['load_demand_kwh', 'solar_output_kw', 'wind_output_kw']].values.astype(np.float32) inputs = torch.tensor(features) with torch.no_grad(): outputs = model(inputs).numpy().flatten() # Average score as a simple indicator stability_score = np.mean(outputs) return stability_score # ---------------------------- # Constants # ---------------------------- API_KEY = "84e26811a314599e940f343b4d5894a7" DEFAULT_LOCATION = "New York" # Sidebar st.sidebar.title("Real-Time Smart Grid Dashboard") location = st.sidebar.text_input("Enter Location", DEFAULT_LOCATION) # Fetch and display weather data weather = fetch_weather(API_KEY, location) if weather: st.sidebar.write(f"Temperature: {weather['temperature']} °C") st.sidebar.write(f"Wind Speed: {weather['wind_speed']} m/s") st.sidebar.write(f"Weather: {weather['weather']}") # Main dashboard title st.title("AI Powered Smart Grid Dashboard for Integration Stability and Energy Trade") # User upload section uploaded_file = st.file_uploader("Upload your Smart Grid CSV file", type=["csv"]) if uploaded_file is not None: data = pd.read_csv(uploaded_file) st.success("File uploaded successfully!") else: st.info("No file uploaded. Using synthetic data for demonstration.") data = generate_synthetic_data() # Required columns check required_columns = ["timestamp", "load_demand_kwh", "solar_output_kw", "wind_output_kw", "grid_health"] if all(col in data.columns for col in required_columns): # Plot load demand fig_load = px.line(data, x="timestamp", y="load_demand_kwh", title="Load Demand Over Time") st.plotly_chart(fig_load) # Renewable energy contribution plot fig_renew = px.bar( data, x="timestamp", y=["solar_output_kw", "wind_output_kw"], title="Renewable Energy Contributions", labels={"value": "Power (kW)", "variable": "Energy Source"} ) st.plotly_chart(fig_renew) # Grid health overview st.subheader("Grid Health Overview") grid_health_counts = data["grid_health"].value_counts() st.bar_chart(grid_health_counts) # PyTorch model prediction - Stability score stability_score = predict_stability(data) st.subheader("Predicted Grid Stability Score (0=Low, 1=High)") st.metric(label="Grid Stability Score", value=f"{stability_score:.3f}") # Optimization recommendations based on last timestamp current_demand = data["load_demand_kwh"].iloc[-1] current_solar = data["solar_output_kw"].iloc[-1] current_wind = data["wind_output_kw"].iloc[-1] recommendation = optimize_load(current_demand, current_solar, current_wind) st.subheader("Recommendations") st.write(f"Current Load Demand: {current_demand} kWh") st.write(f"Solar Output: {current_solar} kW") st.write(f"Wind Output: {current_wind} kW") st.write(f"Recommendation: {recommendation}") else: st.error(f"The uploaded data is missing one or more required columns: {required_columns}")