cmapss-dataset

This repository provides a framework for Federated Prognostics and Health Management (PHM) using Temporal Convolutional Networks (TCN) to predict the Remaining Useful Life (RUL) of industrial assets. Using the NASA CMAPSS dataset, it enables comparative studies between centralized training baselines and federated learning environments.

基于时空图神经网络的航空发动机剩余寿命（RUL）预测 | STGNN-based RUL prediction using NASA C-MAPSS

基于时空图神经网络的航空发动机剩余寿命（RUL）预测 | STGNN-based RUL prediction using NASA C-MAPSS

Native C++ predictive maintenance application for NASA C-MAPSS FD001 turbofan data, featuring RUL estimation, sensor diagnostics, fleet analysis, model comparison, and a Qt-based desktop viewer.

End-to-end predictive maintenance: predicts jet engine Remaining Useful Life (RUL) on NASA C-MAPSS FD004

GaugeNet: geometry-aware unified RUL prediction on NASA C-MAPSS. A single 746K-parameter causal-attention model, unchanged across FD001-FD004, using Riemannian and Wasserstein drift from a healthy reference as input. RMSE 9.05-10.93, best NASA Score on FD002/FD004. Trained weights and Gradio demo.