1. Data Source
All data in this repository were generated using the Gym Electric Motor (GEM) simulation platform:
🔗 https://github.com/upb-lea/gym-electric-motor

The platform provides a baseline Model Predictive Control (MPC) implementation. We modified this code to simulate current, torque, and angular velocity control tasks for permanent magnet synchronous motors (PMSMs) under varying configurations.


--------------------------------------------------------------------------------------------

2. Dataset Structure
This data package is divided into two parts:

part_1_simulation_data/
├── current_control/        # 50 samples (motor configurations) under current control
├── speed_control/          # 50 samples under speed control
└── torque_control/         # 50 samples under torque control


Each folder contains time-series simulation data for one control mode across 50 motor instances with different parameters.

Part 2: Derived Data with Damping Coefficient Perturbation
This section includes datasets derived from the original simulations by modifying the angular velocity dynamics. A damping coefficient B was added to the equation to simulate parameter drift and system evolution.

part_2_damping_modified_data/
├── new_omega_data/              # Updated omega data with damping coefficients
│   ├── b_list.txt               # List of all used damping coefficient values
│   └── sc_b.py                  # Script for generating damping coefficients
│
├── sim_data_XXX.pkl             # Base motor parameter files (XXX = 0–9, 10 motors total)
├── b_list.txt                   # Output of sc_b.py
└── get_motor_para.py            # Script to extract parameters from .pkl files


The new_omega_data/ directory contains time-series data of angular velocity under different damping conditions, simulating adaptive scenarios. Each motor was tested across 4 damping intervals with 50 values per interval, totaling (50 original + 200 derived) × 10 motors.



3. Reproduction Notes
Simulation was executed using Python 3.8 and Gym Electric Motor 1.1.

For reproducibility:

Start from Part 1 to load the base dataset.

Use get_motor_para.py to extract motor parameters from .pkl files.

Use sc_b.py to generate damping coefficient variations (relying on the list in b_list.txt).

All values are stored in NumPy or Pickle format. File-specific details can be found as inline comments in the respective .py files.