Dataset for Validation of Tans-Layer Model Learning of Multi-component Systems

Published: 13-05-2018| Version 4 | DOI: 10.17632/cmd3z85t56.4
Zhixue Tan,
Shisheng Zhong,
Lin Lin


Enclose are two datasets: The first dataset is a industrial dataset generated by a commercial twin-shaft turbofan engine, which produced about 33000 lbs thrust for aircrafts. This data set contains 500 samples of the thermal-dynamical reading inside the engine, which characterized the profile of the steady cruise state of this type of engine. This dataset was meant to validate the modeling accuracy and efficiency of Trans-Layer Model Learning ( TLML). The second dataset is 10 collections of system output of a simulative multi-component systems containing a paralell structure and a curcuit for closed-loop control. each collection contains 500 samples dispersed across the operation regime of the system, this dataset is meant to testify the capability of TLML to overcome measurement dificiency and generate high-fidelity models for components;


Steps to reproduce

Both datasets are unique and unreproducable.