Thermal Field Prediction on Lunar Rover Thermal Simulation (test)
0.0033Average RMSEPIML
Evaluation results
| Method | Links | |
|---|---|---|
| PIMLNumber of Hidden Layers=5, Nodes Per Layer=1024, Learning Rate=10−3, Number of Training Samples=1000, Number of Testing Samples=200, Number of Epochs=500, Activation Function=Leaky ReLU, Number of Outputs=522026.05 | 0.0033 | |
| ANNNumber of Hidden Layers=5, Nodes Per Layer=1024, Learning Rate=10−3, Number of Training Samples=1000, Number of Testing Samples=200, Number of Epochs=500, Activation Function=Leaky ReLU, Number of Outputs=147682026.05 | 0.0054 | |
| LFModel Type=Low-Fidelity physics model2026.05 | 0.0065 |