An intelligent scheme for rapid INS alignment procedure using artificial neural networks

Yun Wen Huang, Yu Sheng Huang, Kai Wei Chiang

Research output: Contribution to conferencePaperpeer-review

Abstract

Inertial navigation systems are commonly used in several applications such as aerospace systems and land vehicle navigation. The navigation parameters including position, velocity and attitude of a moving platform are determined by processing the measurement of inertial sensors. In general, the accuracy of the navigation solutions provided by an INS depends on the initial attitude angles of the body frame where the measurements of specific forces and angular rate are sensed by the inertial measurement unit and the navigation frame applied. Therefore, those initial angles have to be estimated accurately prior to switching the INS into navigation mode. The techniques to estimate those initial attitude angles are known as the I process of alignment. An optimal estimator, the Kalman filter, takes about 10 to 15 minutes to converge then achieve the alignment process due to measurement errors. Those errors increase the alignment time and deteriorate the overall accuracy of initial attitude angels estimated. Therefore, this article suggested an intelligent alignment scheme that combines an Artificial neural network and Kalman filter to improve the accuracy of initial attitude angles and reduce the consumption of time. In this study, a navigation grade inertial measurement unit was applied to verify the performance of proposed scheme. The preliminary results presented in this article indicate that a faster alignment procedure with higher accuracy can be achieved through the use of proposed scheme.

Original languageEnglish
Pages320-326
Number of pages7
Publication statusPublished - 2007 Aug 23
EventInstitute of Navigation National Technical Meeting, NTM 2007 - San Diego, CA, United States
Duration: 2007 Jan 222007 Jan 24

Other

OtherInstitute of Navigation National Technical Meeting, NTM 2007
CountryUnited States
CitySan Diego, CA
Period07-01-2207-01-24

All Science Journal Classification (ASJC) codes

  • Engineering(all)

Fingerprint Dive into the research topics of 'An intelligent scheme for rapid INS alignment procedure using artificial neural networks'. Together they form a unique fingerprint.

Cite this