Abstract | ||
---|---|---|
In the complex environment of high dynamic and strong interference, gyros strapdown to the high spinning flying body are affected by high rotation and high overload. As a result, it is impossible to accurately obtain the angular rate information of high spinning flying body. In this article, a virtual gyros construction method based on deep learning is proposed. In this method, the physical model of virtual gyros is constructed according to the motion characteristics of high spinning flying body and the characteristics of magnetoresistive sensors and accelerometers output data. Bidirectional long short-term memory (BILSTM) is introduced to predict and solve the attitude change quaternion for high spinning flying body, and then, virtual gyros can be obtained via the relationship between attitude change quaternion and angular rate. Simulation and experiment results show that virtual gyros physical model is feasible and accurate, and prediction accuracy of BILSTM is better than gated recurrent unit (GRU) and long short-term memory (LSTM). |
Year | DOI | Venue |
---|---|---|
2022 | 10.1109/TIM.2022.3212544 | IEEE Transactions on Instrumentation and Measurement |
Keywords | DocType | Volume |
Accelerometers,bidirectional long short-term memory (BILSTM),high spinning flying body,magnetoresistive sensors,virtual gyros | Journal | 71 |
ISSN | Citations | PageRank |
0018-9456 | 0 | 0.34 |
References | Authors | |
7 | 3 |
Name | Order | Citations | PageRank |
---|---|---|---|
Jinwen Wang | 1 | 0 | 0.68 |
Zhi-Hong Deng | 2 | 185 | 23.33 |
Kai Shen | 3 | 2 | 2.05 |