Skip to content

AKing-283/37_Sustainability

Repository files navigation

Team_name: Snap

Team_id: 37_Sustainability

Hackwave Hackathon

Me and my teammates have participated in a hackathon organised by A.P.Shah Institute of Technology . My team SNAP! @Aking-283. @ShadowAniket @Shreyasbhalekar @navyaagrawal20 got to learn what's machine learning and what one can do by creating the moddels which predicts the future data but takes input as past few years data and gets trained.But during thhe hackathon we got to learn diiferent ML algorithms.Commits to this project are appreciated.

Smart Grid Optimization: Integrating Renewable Energy for Sustainable Power Systems

Introduction

This project tackles the challenge of integrating wind energy into smart grids to create sustainable power system functionalities. The smart grid power generation follows the 4-Node Star Architecture.

Synopsis

This project utilizes an extensive dataset at an hourly level of granularity, including:

Renewable wind energy generation data. Smart grid data indicating stability status. The objectives are:

To build predictive models capable of accurately forecasting power generation, with forecasted values of independent variables provided. To assess grid stability. One generation node has three consumption nodes, requiring the generated power to be divided between the three nodes according to a specified ratio. This approach is designed to empower operators to make real-time, data-driven decisions. The project involves hands-on data analysis, feature engineering, and model development to understand the complex interplay between smart grid dynamics, renewable energy generation, and overall grid performance.

Problem Statement

Using the past 5 years of data provided, predict the total power generated (p) for the first 3 months of 2024 at an hourly granularity. The independent variables for this period are provided in a separate file. Merge all 20 data files into a final input file containing DateTime, Air Pressure, Wind Speed, and Power Generated variables. Based on the predicted total power generated (p), distribute the power to 3 different consumption nodes (refer to the 4-Node Star Architecture) with Node 1 receiving 20%, Node 2 receiving 45%, and Node 3 receiving 35% of the total power generated. Use the existing and newly prepared dataset to determine the Stability (“stability,” Dependent Variable) of the power grid. For the above problem, consider the Grid folder data, which includes price per unit, unit consumption, and grid stability reports on an hourly basis. Add the power generated and stored at each node (predicted in step 3). The final report can be presented in different ways: Analyze the output to determine the percentage of 'Stable' and 'Unstable' grid conditions over the three-month span. This analysis will provide valuable insights into the overall stability trends of the grid. Additionally, identify patterns in the hours when the grid is most prone to instability, and examine the frequency and duration of 'Unstable' conditions throughout the day. These insights can inform grid operators about potential vulnerabilities and enable proactive measures to mitigate instability risks during critical hours. Note: Insights are not limited to these examples. Additional views and analyses are encouraged.

Data Description

The dataset includes 5 years of data on three parameters to be used to determine power generated by wind energy. The dataset is divided by year, with each parameter in a separate file. The parameters are:

Air Temperature (°C) Pressure (atm) Wind Speed (m/s) Power Generated (MW) [Target Variable] Additionally, the smart grid dataset encompasses various parameters:

Power Generated (MW): power_gen_1, power_gen_2, power_gen_3 (total power generated is distributed into each node as 20%, 45%, & 35%, respectively) Price per unit: p1, p2, p3 Power Consumption: C1, C2, C3 (arithmetically negative as it is being consumed) Stability: A categorical (binary) label ('stable' or 'unstable') [Target Variable] Note:

All independent variable forecasted values for January, February, and March 2024 are provided in a file. Target variables for model validation over the three-month period are also included. All datasets are timestamped at an hourly granularity, providing insights into the temporal dynamics of energy generation and consumption within the smart grid environment. Dataset Source Dataset Link: https://drive.google.com/drive/folders/1trAEOsroNR4TDKDp1aTk-C7F1RwhoD15?usp=drive_link

#Impact The aim of this project is to foster innovation in sustainable energy systems by developing machine learning solutions for smart grid optimization. Through data-driven decision-making, the project aims to contribute to the advancement of grid reliability, resilience, and efficiency. The insights gained from this project are intended to inform policymakers, industry stakeholders, and research communities about the opportunities and challenges associated with renewable energy integration, ultimately accelerating the transition toward a low-carbon energy landscape.

This version is tailored to a team that might be new to hackathons, emphasizing the project work and objectives without assuming prior hackathon experience.This version may seem to be incomplete and further commits are appreciated which will continue this project.

If any issue occurs just contact us.

About

Hackwave Hackathon

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published