← Latest papers
⚡ electrical engineering

Power System Studies Using Open-Access Software

This paper proposes a flexible simulation environment using R/RStudio as a control center to bidirectionally integrate open-access tools like EMTP/ATP and OpenDSS, thereby expanding their capabilities for steady-state and transient power system studies and enabling the application of machine learning algorithms for performance prediction.

Original authors: Juan A. Martinez-Velasco, Pau Casals-Torrens, Ricard Bosch-Tous, Alexandre Serrano-Fontova

Published 2026-03-25
📖 5 min read🧠 Deep dive

Original authors: Juan A. Martinez-Velasco, Pau Casals-Torrens, Ricard Bosch-Tous, Alexandre Serrano-Fontova

Original paper licensed under CC BY 4.0 (http://creativecommons.org/licenses/by/4.0/). This is an AI-generated explanation of the paper below. It is not written or endorsed by the authors. For technical accuracy, refer to the original paper. Read full disclaimer

Imagine you are trying to predict the weather. You have two powerful tools: a super-accurate weather simulator that tells you exactly how wind and rain will hit a specific mountain, and a giant spreadsheet that can crunch millions of numbers to find patterns in past storms.

This paper is about gluing those two tools together using a "smart conductor" (a software called R/RStudio) to create a super-powerful lab for studying electricity grids.

Here is the breakdown of the paper in simple terms:

1. The Problem: Too Many Tools, Too Much Work

Power engineers usually have to use expensive, specialized software to simulate how electricity behaves.

  • Tool A (ATP): Great for simulating fast, scary events like lightning strikes or sudden power surges. It's like a high-speed crash test dummy for electricity.
  • Tool B (OpenDSS): Great for looking at how electricity flows through a neighborhood over a whole day or year. It's like a traffic manager for a city.
  • The Issue: These tools are great at their jobs, but they are a bit "dumb" when it comes to statistics or spotting complex patterns. They just spit out numbers.

2. The Solution: The "Smart Conductor" (R/RStudio)

The authors propose using R/RStudio as the "brain" or "conductor" of the orchestra.

  • R is a free, open-source tool famous for statistics and drawing beautiful charts.
  • The Setup: Instead of manually typing data into the crash-test tool and then copying results into a spreadsheet, the authors built a system where R talks to the other tools automatically.
    • R tells the tools: "Run a simulation with this specific lightning bolt."
    • The tools run the simulation.
    • The tools send the results back to R.
    • R organizes the data, draws the graphs, and even uses Machine Learning (AI) to guess what will happen next time.

3. The Four "Practice Drills" (Case Studies)

To prove this system works, the authors ran four different experiments. Think of these as training drills for a new robot:

  • Drill 1: The Fault Detective (ATP + R)

    • Scenario: A power line breaks. Where did it break? What kind of break was it?
    • The Trick: They ran thousands of simulations of different breaks. They fed the results into an AI (called k-Nearest Neighbors).
    • Result: The AI learned to look at the voltage numbers and say, "Ah, that looks like a break 30km down the line!" It got almost 100% right.
  • Drill 2: The Lightning Predictor (ATP + R)

    • Scenario: A lightning storm is coming. Will it hit the power lines and cause a blackout?
    • The Trick: They simulated 50,000 random lightning strikes using a "Monte Carlo" method (basically rolling the dice thousands of times to see all possible outcomes).
    • Result: They used an AI (Support Vector Machine) to predict if a specific lightning strike would cause a flashover. The AI was 99% accurate. It's like having a crystal ball that tells you if your house will get struck by lightning based on the cloud shape.
  • Drill 3: The Neighborhood Manager (OpenDSS + R)

    • Scenario: A neighborhood has solar panels and battery packs. How does the power flow change when the sun goes down or when everyone turns on their AC?
    • The Trick: They simulated random loads (people turning appliances on/off) and different battery strategies.
    • Result: They used R to visualize the data and see which battery strategy saved the most energy. It showed that having a battery helps, but the strategy for when to charge/discharge matters even more.
  • Drill 4: The Stability Guardian (OpenDSS + R)

    • Scenario: A huge power plant is running. If a short circuit happens, will the whole grid collapse, or will it bounce back?
    • The Trick: They ran simulations with different fault durations and power levels. They trained three different AI models (Neural Networks, SVM, kNN) to predict the outcome.
    • Result: The AI learned to predict grid stability with 100% accuracy. It's like a coach who watches a thousand practice games and can instantly tell you if the team will win or lose based on the first few seconds of play.

4. Why This Matters (The "So What?")

  • It's Free: All the software used is free and open-source. You don't need to buy expensive licenses.
  • It's Flexible: Because R is the "conductor," you can swap out the tools or change the rules easily.
  • It's Smart: The real magic is Machine Learning. Instead of just simulating one scenario, you can simulate thousands, teach a computer to learn the patterns, and then use that computer to predict the future instantly without running a new simulation every time.
  • It's Visual: R makes it easy to turn boring numbers into clear, colorful charts that anyone can understand.

The Bottom Line

This paper is a recipe for building a super-lab for power grids. By combining specialized simulation tools with a smart statistical brain (R), engineers can run more tests, learn faster, and use AI to predict power problems before they happen—all without spending a fortune on software. It turns a slow, manual process into an automated, intelligent workflow.

Drowning in papers in your field?

Get daily digests of the most novel papers matching your research keywords — with technical summaries, in your language.

Try Digest →