https://doi.org/10.1140/epjp/s13360-024-05916-3
Regular Article
A novel approach for the temperature prediction of ring laser gyroscope using teamwork optimization enabled bias-compensated long short-term memory
1
ISRO Department, ISRO Inertial Systems Unit (IISU), Vattiyoorkavu Complex, Nettayam, 695013, Thiruvananthapuram, Kerala, India
2
ISRO Department, Vikram Sarabhai Space Centre, 695022, Thiruvananthapuram, Kerala, India
3
Physics Department, IIST, Trivandrum, Indian Institute of Space Science and Technology, Valiamala, Thiruvananthapuram, Kerala, India
Received:
24
October
2024
Accepted:
9
December
2024
Published online:
20
December
2024
Ring laser gyro (RLG) has turned out to be the predominant rotation rate sensor used in high-precision strap-down inertial navigation systems (SINS), extensively employed in the aerospace field nowadays. However, in actual space environments, temperature variation within the device leads to an uneven RLG bias error which results in performance degradation of SINS. To resolve this issue, diverse compensation methods have been performed carefully with comprehensive analysis to enhance environmental adaptability. Moreover, a bias of RLG changes with temperature in a nonlinear manner, which is a significant restraining factor to enhance the accuracy of RLG. In this paper, a novel method is developed using a deep learning model to predict the temperature model for the sensor. Initially, sensor data are pre-processed by min–max normalization and then, it is subjected to feature selection using correlation. The temperature prediction is carried out using bias-compensated long short-term memory (BC–LSTM) after selecting significant features, and the teamwork optimization algorithm (TOA) is used as the training algorithm to compensate for the bias error that occurs due to nonlinearity in the data. Moreover, the classification of sensors has been done using the proposed TOA-LSTM to identify the best sensors for critical space applications. The performance of proposed model is analysed using evaluation metrics such as accuracy, True Positive rate (TPR), True Negative rate (TNR), False Positive rate (FPR), mean squared error (MSE), mean absolute error (MAE) and root-mean-squared error (RMSE). The experimental investigation states that the proposed approach attains accuracy, TPR, TNR, and FPR with values of 0.960, 0.966, 0.949 and 0.0511, respectively.
Copyright comment Springer Nature or its licensor (e.g. a society or other partner) holds exclusive rights to this article under a publishing agreement with the author(s) or other rightsholder(s); author self-archiving of the accepted manuscript version of this article is solely governed by the terms of such publishing agreement and applicable law.
© The Author(s), under exclusive licence to Società Italiana di Fisica and Springer-Verlag GmbH Germany, part of Springer Nature 2024
Springer Nature or its licensor (e.g. a society or other partner) holds exclusive rights to this article under a publishing agreement with the author(s) or other rightsholder(s); author self-archiving of the accepted manuscript version of this article is solely governed by the terms of such publishing agreement and applicable law.
