Servomotor-Driven Ballscrew Mechanism Degradation Data Set
Servomotor-Driven Ballscrew Mechanism Degradation Data Set
- Provided by: GE Research and University of Tennessee Knoxville
- Acknowledgement: This dataset and work presented was funded in part by the Advanced Research Projects Agency-Energy (ARPA-E), U.S. Department of Energy, under Award Number DE-AR0001290. The views and opinions of authors expressed herein do not necessarily state or reflect those of the United States Government or any agency thereof.
- Download (21.3 GB): https://phm-datasets.s3.amazonaws.com/GE-UTK/FMCRD_Data.zip
Description: This dataset is a record of the inputs, outputs and annotations of a servomotor simulator designed using MathWorks Simulink. The simulation models servomotor operation applied to cases where the motor operates intermittently. One example of such an application is the fine motion control rod drive (FMCRD) mechanism used for intermittent, and typically infrequent, fine motion (insertion or withdrawal) adjustment of control rods in some nuclear reactor designs for optimizing and shaping core power in the reactor. This is in contrast to most rotating machinery that operates on a more continuous basis and sustains wear over steady state operation. Although intermittently operational drives do not run continuously, the servomotors and associated linear motion mechanisms show wear and damage, due to internal and external causes, during their operational lifetime. This causes impediments to the movement of the rotor shaft. The resultant modes of degradation are different from continually operating servomotors. The prognostics and health management for such an intermittent operational regime is different, and this dataset addresses the need for data and algorithms for such applications. Specifically, for machine learning approaches, feature extraction over periods of transient operation requires different preprocessing and signal analysis than features one may extract over continuously operating machinery.
While the causes of servomotor drive degradation can be diverse, in simulation the impact of cumulative damage is modeled as an external opposing load which resists the movement of the motor shaft. This load is a latent unobservable variable, but its effects may be manifested in observable signals from the Simulink model. The loads are a sampled from a mixture of Gaussians with 4 distinct modes leading to an assessment of degradation represented in the dataset as
- A categorical degradation level called ‘ylabel’ in the dataset which is one of like-new, low, medium or high, abbreviated as {‘LN’, ‘LO’, ‘MED’, ‘HI’}, and
- A floating-point degradation value, denoted ‘DV’ in the dataset.
The dataset is in the form of CSV files, with distinct files for training and testing. Each CSV file contains the following columns which include measurements such as motor currents and rotor speed which would be part of the instrumentation in a typical deployments of rotating machinery. These observable measurements can be used to predict the health state of the servomotor.
Column name | Description |
time | Timestamp for each run |
DV | Degradation value |
rod_demand_pos | Demanded rod position |
rod_actual_pos | Actual rod position |
torque | Motor torque |
rotor_speed | Rotor speed |
i_3p_a | Stator current: Phase A |
i_3p_b | Stator current: Phase B |
i_3p_c | Stator current: Phase C |
direct | Direct component of DQZ transform on the stator currents |
quadrature | Quadrature component of DQZ transform |
run_index | Index of runs |
transitions | Index of transitions |
del_pos | Demanded displacement, i.e., position delta |
ylabel | Categorical degradation levels; ylabel ∈ {“LN”, “LO”, “MED”, “HI”} |