This undergraduate thesis proposes an innovative approach to the control of a motor-generator test bench, integrating advanced optimization, modeling, and control techniques. The study utilizes genetic algorithms (GAs) developed with the PyGad library for offline optimization in tuning Proportional-Integral (PI) and Proportional-Integral-Derivative (PID) controllers. Data collection from the test bench is performed through a specific code, while another code is developed for building the model using the
Sparse Identification of Nonlinear Dynamics (SINDy) technique. This model serves as a foundation for the development of controllers. Two sets of codes are implemented for PI and PID controllers. The first set, of an offline nature, employs GAs with PyGad for optimizing controller parameters. The second set is online and functions to transmit the controller obtained through optimization to the motor-generator test bench in real-time. The study addresses theoretical and practical aspects, providing an in-depth analysis of the results obtained with the implementation of PI and PID controllers, comparing the performance of the two tuning methods. Additionally, significant contributions are presented in the context of dynamic systems control, exploring the effectiveness of integrating modern techniques for optimization and modeling.
|
