Sistema de Predicción Energética con IA

Sistema de Predicción Energética con IA is an MCP server that provides domestic energy consumption forecasting and analysis directly within Model Context Protocol clients like Claude AI. Developed as an academic Master's thesis project, it connects machine learning ensemble models (built with Scikit-learn, XGBoost, and LightGBM) to local household energy data and external electricity price APIs from ESIOS (Red Eléctrica de España). Data analysts, homeowners, and researchers use it to inspect historical consumption metrics, generate forward-looking appliance-level usage predictions, and optimize electricity schedules based on pricing tiers. The system also integrates SHAP (SHapley Additive exPlanations) values to explain which features drive specific consumption predictions. By exposing dedicated tools like predict_consumption, get_consumption_analysis, and get_precio through an MCP interface, the server enables conversational agents to evaluate power demand trends, query live and fallback energy rates, and suggest practical energy optimization steps.

Category: Data & Analytics

Tags: analytics, consumption, energy, forecasting, sustainability

Visit Sistema de Predicción Energética con IA

How to install and configure Sistema de Predicción Energética con IA

1. Ensure you have Python 3.8 or higher, an active ESIOS API token, and a tool runner like uv installed. 2. Clone the repository and navigate to the project directory: git clone https://github.com/Davad122/TFM. 3. Open your Claude AI or MCP client configuration file. 4. Register the server under mcpServers using the provided configuration: json { "mcpServers": { "mcp-david-TFM": { "command": "uv", "args": [ "--directory", "RUTA_DEL_PROYECTO", "run", "-m", "davidElectric" ], "env": { "ESIOS_API_TOKEN": "YOUR_ESIOS_TOKEN" } } } } 5. Replace RUTA_DEL_PROYECTO with your absolute project path and insert your actual ESIOS API key into ESIOS_API_TOKEN. 6. Restart your MCP client to enable the energy forecasting tools.

What you can do with Sistema de Predicción Energética con IA

• Forecasting future domestic power consumption across specified date ranges using predict_consumption. - Reviewing historical appliance energy usage breakdowns and demand trends via get_consumption_analysis. - Analyzing machine learning prediction factors for specific home appliances using SHAP values with explain_predictions. - Querying current and historical Spanish electricity tariffs via the ESIOS API using get_precio. - Finding optimal energy costs using get_precio_inteligente to automatically handle fallback pricing for targeted dates.

Key facts

• https://github.com/Davad122/TFM
• Data & Analytics, Maps, Weather & Local Data
• analytics, consumption, energy, forecasting, sustainability

Part of MCP Servers

What is Sistema de Predicción Energética con IA?

It is an academic machine learning system and MCP server designed to analyze domestic energy consumption, predict future demand, explain model outputs with SHAP values, and query Spanish electricity prices via the ESIOS API.

Which MCP clients work with this server?

The server is built to run through standard Model Context Protocol implementations, such as Claude AI desktop or any client that supports standard I/O MCP server configurations.

What tools are available via this MCP server?

The server exposes five main tools: predict_consumption for usage forecasting, get_consumption_analysis for historical appliance data, explain_predictions for SHAP interpretability, get_precio for electricity rates, and get_precio_inteligente for pricing with automatic fallback.

Is Sistema de Predicción Energética con IA open source?

The project is licensed under an academic license for educational evaluation, reference, and non-commercial study with attribution. Commercial usage, redistribution, and unauthorized modification are strictly prohibited by the author.