Basic report added. Report missing significant number of graphs and conclusions
@ -0,0 +1,5 @@
|
||||
# Inverted Pendulum Goal
|
||||
The goal of the inverted pendulum project is to explore different loss functions and how they affect the training and performance of
|
||||
physics informed neural networks for model predictive control. The project specifically focuses on varying the base loss function from mean squared error (MSE) to different polynomials as well as applying weighting functions to the standard MSE loss function.
|
||||
|
||||
The current report file is [here](report/Physics_Informed_Neural_Network_Controller_Training_for_Inverted_Pendulum_Systems_Using_Variable_Base_Loss_Functions_and_Time_Weighted_Loss_Functions.pdf)
|
||||
|
Before Width: | Height: | Size: 268 KiB After Width: | Height: | Size: 266 KiB |
|
Before Width: | Height: | Size: 268 KiB After Width: | Height: | Size: 284 KiB |
|
Before Width: | Height: | Size: 280 KiB After Width: | Height: | Size: 292 KiB |
|
Before Width: | Height: | Size: 265 KiB After Width: | Height: | Size: 270 KiB |
|
Before Width: | Height: | Size: 275 KiB After Width: | Height: | Size: 282 KiB |