Skip to content
Snippets Groups Projects
Commit cfdffb80 authored by Daniel Jung's avatar Daniel Jung
Browse files

updated details regarding training data

parent f079da0e
No related branches found
No related tags found
No related merge requests found
Pipeline #163159 passed
......@@ -76,7 +76,7 @@ The considered known faults in the competition are:
The sensor faults are introduced in the engine control unit by multiplying the measured signal x with a factor y = (1 + f)*x to get the sensor output y. A fault means that the factor f is not equal to 0. Note that the magnitude f could be both positive and negative. The sensor faults can affect the operation of the system, and not only the sensor output, due to feedback loops. The leakage is induced by opening a valve connected to the intake manifold.
Training data consists of eight datasets. The dataset contains one to two datasets from each known fault where each dataset contains one realization (magnitude) of each fault. Note that the training data is not representative of all fault magnitudes and operating conditions of the system. Thus, the diagnosis system should be able to handle faults of different magnitudes. The naming of each dataset is done as follows:
Training data consists of 26 datasets. The data contains two to seven datasets from each known fault where each dataset contains one realization (magnitude) of each fault. Note that the training data is not representative of all fault magnitudes and operating conditions of the system. Thus, the diagnosis system should be able to handle faults of different magnitudes. The naming of each dataset is done as follows:
'driving_cycle'\_'fault'\_'fault magnitude'.csv
......
0% Loading or .
You are about to add 0 people to the discussion. Proceed with caution.
Please register or to comment