MRLocoSet

MRLocoSet: Simulation and experimental dataset for magnetic microroller locomotion, including state/control variables and dynamic linearization terms (e.g., φ matrices) in obstacle environments.

Paper status: The associated manuscript is currently under preparation. Paper information will be added in a future update.

Highlights

• Simulation + Experiment: both simulated and real microroller locomotion data are provided.
• Data-driven modeling oriented: each file includes state/control variables and dynamic linearization-related terms (e.g., φ matrix elements).
• Obstacle environments: trajectories are collected in obstacle fields similar to the example shown below. Obstacle environment example If the image is not displayed, please check that it exists at: docs/figures/obstacle_environment.png

Data Statistics

• Simulation: 6,279 trajectories (.txt), packaged into 7 zipped batches.
• Experiment: 82 trajectories (.txt), stored as plain text files.

File Format (per .txt file)

Each .txt file is a comma-separated text file with a header row. A typical header looks like: sequence, delta(k-1), f(k-1), theta(k), v(k), delta_diff(k-1), f_diff(k-1), theta_diff(k), v_diff(k), phi(1,1;k), phi(1,2;k), phi(2,1;k), phi(2,2;k)

• sequence: sample index
• delta, f, theta, v: state/control-related variables used for microroller locomotion modeling
• *_diff: difference terms
• phi(·,·;k): elements of the dynamic linearization matrix at time step k