/PLASMAG

Primary LanguagePythonGNU General Public License v3.0GPL-3.0

PLASMAG

(Python Library for Accurate Space Magnetometer Adjustments with Genetics)

PLASMAG is a simulation software specifically designed for space magnetometers. At its core, PLASMAG serves as a comprehensive tool for the parameters adjustment.

Currently, PLASMAG is tailored to support search coil type instruments. However, its architecture is built with flexibility and extensibility in mind. This means PLASMAG is not only limited to current implementations but is also designed to easily accommodate the addition of new instruments and the integration of diverse calculation methods and models.

Online documentation can be found here

Table of Contents

Quick Setup

Clone the repository :

git clone https://github.com/LaboratoryOfPlasmaPhysics/PLASMAG

Installation

Prerequisites

Ensure you have Conda installed. If not, follow the instructions here: https://conda.io/projects/conda/en/latest/user-guide/install/index.html

Linux conda installation

Linux Conda Installation

Install Conda using the following commands:

mkdir -p ~/miniconda3
wget https://repo.anaconda.com/miniconda/Miniconda3-latest-Linux-x86_64.sh -O ~/miniconda3/miniconda.sh
bash ~/miniconda3/miniconda.sh -b -u -p ~/miniconda3
rm -rf ~/miniconda3/miniconda.sh

Add conda to your PATH

~/miniconda3/bin/conda init bash
~/miniconda3/bin/conda init zsh

Restart your terminal and verify the Conda installation:

conda --version

Setting Up the Project

Create and activate a new Conda environment:

conda env create -f environment.yml

Activate the environment

conda activate PLASMAG

Run the application

python PLASMAG.py

Usage

  1. Simulation Capabilities:
  • Full Electrokinetic Simulation: Uses SPICE kernel via PySPICE for comprehensive electrokinetic modeling.
  • (WIP) Real Data Fitting: Incorporates a two-step data fitting tool that includes denoising and neural network fitting for accurate simulations based on measured data.
  1. Visualization Tools:
  • Adaptive Input Section: Adjusts input options based on the magnetometer parameter set.
  • Real-time Parameter Variation: Sliders and real-time calculations to visualize the impact of parameter changes.
  • Modular Plot Section: Allows up to 5 simultaneous displays of various products, providing a comprehensive view of simulation results.
  1. Modularity and Flexibility:
  • Strategy Design Pattern: Implements a modular engine design using the strategy pattern, allowing dynamic changes in computation methods and easy adaptation to new requirements.
  • Parallel Computation: Supports parallelization of computations for improved efficiency, to be implemented for futures improvements.
  • Graph Visualization Tools: Provides full graph visualization to monitor computation dependencies and processes.
  1. Export and Import Capabilities:
  • Export Data: Allows exporting of any product or parameter for plotting or later re-importation.
  • NEMI/CLTF Export: Exports the NEMI and Closed Loop Transfer Function (CLTF) graphs for easy design sharing.
  1. Optimization Module (WIP):
  • Genetic Optimization (DEAP): Utilizes evolutionary algorithms to find optimal parameter sets.
  • Particle Swarm Optimization: Implements swarm intelligence techniques to explore solution spaces.
  • Simulated Annealing: Employs probabilistic techniques to approximate global optimization.
  • Mono-parameter and Multi-parameter Optimization: Supports optimization for single and multiple parameters.
  • Mono-criteria and Multi-criteria Optimization: Handles single and multiple criteria, combining them into a single framework using weighted polynomials.

Available Targets:

  • Impedance (Resonance Frequency)
  • NEMI (WIP)

Availables Models

At the moment, the PLASMAG simulation model has only one model implemented: a simple search coil MAGNETOMETER analytic model.

The simulation was validated by comparing the simulation results with the real data from the JUICE mission.

The parameters for JUICE's search coil can be found in the data/JUICE.json file."

The implemented simulation model performs quite well, demonstrating high accuracy for low frequency values. This is reflected in the impedance plot, where the plotted curves all maintain consistent levels and shapes.

However, challenges arise when extending the analysis to higher frequency ranges. Specifically, for the NEMI and Closed Loop Transfer Function (CLTF), the model tends to diverge slightly at high frequencies. This divergence becomes particularly evident after the resonance frequency. At this juncture, analytically describing the 'capacitance' part of the system grows increasingly difficult, leading to a slight deviation from expected behaviors. This discrepancy underlines a current limitation of the model, spotlighting the need for further refinement and development to enhance its accuracy.

The model (combination of all calculation strategy) is composed of 2 parts:

  • data/JUICE.json : Contains the parameters of the JUICE search coil magnetometer, their values will be used for computations
  •

Documentation

To generate project documentation:

pip install sphinx
cd docs
make html

Navigate to docs/_build/html/index.html to view the documentation.

Contributing

The project is open to contributions. Please refer to the CONTRIBUTING.md file for more information.

Acknowledgments

  • Maxime RONCERAY - Developer - LPP/CNRS/X contact-me
  • Malik MANSOUR - Supervisor - LPP/CNRS/X malik.mansour@lpp.polytechnique.fr
  • Claire REVILLET - IT/Dev supervisor - CNRS Orleans/LPC2E
  • Guillaume JANNET - Electronics Engineer - CNRS Orleans/LPC2E