DESIGN AND IMPLEMENTATION OF NON LINEAR CONTROLLER WITH PHYSICS INFORMED NEURAL NETWORK APPROACH FOR GANTRY CRANE PROTOTYPE
The cargo handling and logistics industry experiences a high rate of activity. Gantry cranes are essential instruments in the container loading and unloading processes at ports. However, gantry crane control often faces challenges when dealing with changes in system parameters and unexpected environ...
Saved in:
| Main Author: | |
|---|---|
| Format: | Final Project |
| Language: | Indonesia |
| Online Access: | https://digilib.itb.ac.id/gdl/view/83873 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| Institution: | Institut Teknologi Bandung |
| Language: | Indonesia |
| Summary: | The cargo handling and logistics industry experiences a high rate of activity. Gantry cranes are essential instruments in the container loading and unloading processes at ports. However, gantry crane control often faces challenges when dealing with changes in system parameters and unexpected environmental disturbances. This research examines the application of non-linear control using the Physics Informed Neural Network (PINN) method, where the control law model is created using the Interconnection Damping Assignment-Passivity Based Control (IDA-PBC) algorithm. This control system was tested through simulation and direct implementation on a gantry crane prototype. The Proportional Integral Derivative (PID) controller was used as a performance comparison for the control system. The result of this research shows that PINN controller have advantage in minimizing sway with root mean square 1,9 degree compared to PID with rms sway 2,6 degree.
Keywords: PINN IDA-PBC, Non-linear control, gantry crane
|
|---|