2026-10-14T14:10:58Z https://www.scidb.cn/oai

10.11922/sciencedb.00295 2020-11-11T20:55:57Z

2020-11-11 S500 UAV Fight log data，Fight data after Parsing And Source code of Long Short Term Memory Neural Network based Robot Operation Status Monitoring and Fault Identification and Isolation Method doi:10.11922/sciencedb.00295 en The original data in Flydata and Flydata2 are exactly the same. Flydata is processed by csvs2excel.py. Flydata2 is processed by csvs2excel2.py. The codes of csvs2excel.py and csvs2excel2.py are basically the same, but they are different in GPS data processing. csvs2excel2.py performs a linear transformation on the range of GPS data to make the data range smaller. In Flydata and Flydata2 files, each folder represents a flight data. Those beginning with train are training data, and those beginning with test are test data. The fault data is the fault data. Details： lstm_uav_state_check.py Used to train the robot running state monitoring model Already have a trained model：model_use_uyx_pre_x.h5 lstm_uav_state_check_change_data.py Used for testing lstm_uav_state_check_use_model.py lstm_uav_state_check_use_model_fault.py（Substitute fault data into the model） lstm_uav_state_check_use_model_normal.py（Substitute normal data into the model）Substitute different data to verify the model lstm_uav_output_check.py GPS sensor verification module used to train the sensor cross-diagnostic network model Already have a trained model：model_use_y_pre_gps.h5 lstm_uav_output_check2.py For testing lstm_uav_output_check_use_model_fault （Substitute fault data into the model） lstm_uav_output_check_use_model_normal（Substitute normal data into the model）Substitute different data to verify the model - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - Flydata 和 Flydata2 里原始数据完全一致。Flydata 是由csvs2excel.py处理得到的。Flydata2 是由csvs2excel2.py处理得到的。csvs2excel.py与csvs2excel2.py在代码上基本一致，只是在GPS数据处理上有所不同csvs2excel2.py对GPS数据的范围进行了线性变换，使得数据范围更小而已。Flydata 和 Flydata2文件中，每一个文件夹代表一次飞行数据。以train开头的，是训练数据，以test开头的是测试数据。带有fault的是故障数据。 明细栏： lstm_uav_state_check.py 用于训练机器人运行状态监测模型 已经有了训练好的模型：model_use_uyx_pre_x.h5 lstm_uav_state_check_change_data.py 用于测试的文件 lstm_uav_state_check_use_model.py lstm_uav_state_check_use_model_fault.py（故障数据代入模型） lstm_uav_state_check_use_model_normal.py（正常数据代入模型）代入不同数据进行验证模型的文件 lstm_uav_output_check.py 用于训练传感器交叉诊断网络模型中GPS传感器验证模块 已经有了训练好的模型：model_use_y_pre_gps.h5 lstm_uav_output_check2.py 用于测试 lstm_uav_output_check_use_model_fault （故障数据代入模型） lstm_uav_output_check_use_model_normal（正常数据代入模型）代入不同数据进行验证模型的文件 UAV fight log data; Robot Operation Status Monitoring source code; Fault Identification and Isolation Method source code Haibo Xu Liu Li Peixuan Li PUBLIC https://creativecommons.org/licenses/by/4.0/ dataset Science Data Bank