Measurement Simplification

This repo implements the code for Measurement Simplification in ρ-POMDP with Performance Guarantees, now accepted to IEEE T-RO

Installation

Clone the repository:

git clone https://github.com/tomyot/measurement-simplification.git

Install required packages:

pip install -r requirements.txt

Usage

python m_simplification.py

Configuration

The behavior of the belief space planning system can be configured via the config.py file. Parameters such as the scenario, number of landmarks, map size, prior mapping, goal location, and number of paths can be adjusted according to specific requirements.

Class: MeasurementSimplification

Attributes:

• prior: The prior belief state.
• actions_random: A list of random actions.
• actions: A list of predefined actions.
• scenario: The scenario for generating landmarks.
• num_landmarks: The number of landmarks to generate.
• map_size: The size of the map.
• landmarks: A dictionary containing information about the landmarks.
• belief: The GaussianBelief object representing the robot's belief.
• fig_num: The figure number for plotting.

Citation

If you use, compare with, or refer to this work, please cite:

@misc{yotam2023measurement,
      title={Measurement Simplification in \rho-POMDP with Performance Guarantees}, 
      author={Tom Yotam and Vadim Indelman},
      year={2023},
      eprint={2309.10701},
      archivePrefix={arXiv},
      primaryClass={cs.AI}
}