Novel temperature modeling and compensation method for bias of ring laser gyroscope based on least-squares support vector machine  

作　　者：于旭东 王宇 魏国 张鹏飞 龙兴武 

机构地区：[1]College of Opoelectric Science and Engineering, National University of Defense Technology [2]School of Mechanical Engineering,Nanjing University of Science and Technology [3]College of Opoelectric Science and Engineering,National University of Defense Technology

出　　处：《Chinese Optics Letters》2011年第5期37-40,共4页中国光学快报（英文版）

摘　　要：Bias of ring-laser-gyroscope （RLG） changes with temperature in a nonlinear way. This is an important restraining factor for improving the accuracy of RLG. Considering the limitations of least-squares regression and neural network, we propose a new method of temperature compensation of RLG bias building function regression model using least-squares support vector machine （LS-SVM）. Static and dynamic temperature experiments of RLG bias are carried out to validate the effectiveness of the proposed method. Moreover, the traditional least-squares regression method is compared with the LS-SVM-based method. The results show the maximum error of RLG bias drops by almost two orders of magnitude after static temperature compensation, while bias stability of RLG improves by one order of magnitude after dynamic temperature compensation. Thus, the proposed method reduces the influence of temperature variation on the bias of the RLG effectively and improves the accuracy of the gyro scope considerably.Bias of ring-laser-gyroscope （RLG） changes with temperature in a nonlinear way. This is an important restraining factor for improving the accuracy of RLG. Considering the limitations of least-squares regression and neural network, we propose a new method of temperature compensation of RLG bias building function regression model using least-squares support vector machine （LS-SVM）. Static and dynamic temperature experiments of RLG bias are carried out to validate the effectiveness of the proposed method. Moreover, the traditional least-squares regression method is compared with the LS-SVM-based method. The results show the maximum error of RLG bias drops by almost two orders of magnitude after static temperature compensation, while bias stability of RLG improves by one order of magnitude after dynamic temperature compensation. Thus, the proposed method reduces the influence of temperature variation on the bias of the RLG effectively and improves the accuracy of the gyro scope considerably.

关 键 词：Error compensation GYROSCOPES Neural networks Regression analysis Ring lasers Support vector machines Temperature distribution 

分 类 号：TN248[电子电信—物理电子学]