3 DOF Kinova Experimental and Simulation Dataset

Instructions for Using the Kinova 3 DOF Sensory and Simulation Dataset

1. Dataset Structure:

   - Sensory Data Folder:

     - Contains real-time measurements from the Kinova robotic arm.

     - Files are organized by joint data (positions, velocities, torques).

     - Data is stored in XLSX format, with each file corresponding to a specific experiment or time interval.

   - Simulation Data Folder:

     - Includes simulated kinematic and dynamic data of the Kinova arm.

     - Data is stored in XLSX or MAT files, depending on the simulation environment used.

     - Files are categorized based on the type of simulation scenario.

2. Loading the Data:

   - Python:

     - Use the `pandas` library to load XLSX files:

       ```python

       import pandas as pd

       sensory_data = pd.read_excel('path_to_sensory_data.xlsx')

       ```

     - For MAT files, use `scipy.io`:

       ```python

       from scipy.io import loadmat

       simulation_data = loadmat('path_to_simulation_data.mat')

       ```

   - MATLAB:

     - Load XLSX files using `readtable` or `xlsread`:

       ```matlab

       sensory_data = readtable('path_to_sensory_data.xlsx');

       ```

     - Load MAT files directly:

       ```matlab

       load('path_to_simulation_data.mat');

       ```

3. Interpreting the Data:

   - Sensory Data:

     - Columns typically represent joint positions, velocities, and torques.

     - Time stamps (if included) provide temporal context for each measurement.

   - Simulation Data:

     - Simulated joint angles, velocities, and forces are provided.

     - Data is organized to mirror the structure of sensory data for easy comparison.

4. Using the Data for Analysis:

   - Benchmarking:

     - Compare real and simulated data to evaluate the accuracy of simulations.

   - Control Algorithm Development:

     - Use the dataset to train and test control algorithms for robotic arms.

   - Machine Learning:

     - Apply machine learning techniques to predict arm behavior or optimize control strategies based on the dataset.