Autonomous Underwater Vehicle Fault Diagnosis Dataset
Description
The dataset contains 1225 data samples for 5 \textit{fault types} (labels). We divided the dataset into the training set and the test set through random stratified sampling. The test set accounted for $20\%$ of the total dataset. Our experimental subject is `Haizhe', which is a small quadrotor AUV developed in the laboratory. For each \textit{fault type}, `Haizhe' was tested several times. For each time, `Haizhe' ran the same program and sailed underwater for 10-20 seconds to ensure that \textit{state data} was long enough. The \textit{state data} recorded in each test were then used as a data sample, and the corresponding \textit{fault type} was the true label of the data sample. The dataset was used to validate a model-free fault diagnosis method proposed in our paper published in Ocean Engineering（Model-free fault diagnosis for autonomous underwater vehicles using sequence convolutional neural network, Ocean Engineering. 232(2021)108874. https://doi.org/10.1016/j.oceaneng.2021.108874）.