Título: Low Cost INS/GPS Navigation System Integrated by An Adaptive Neural Network Training Kalman Filtering Methodology
Autores: Marques Filho, Edmundo A.; Rios-Neto, Atair; Kuga, Helio Koiti
Resumo: This paper presents the development and testing of an alternative approach to make it feasible to use low cost inertial measurements units (IMU) in integrated Inertial Navigation System (INS) and Global Positioning System (GPS) applied to positioning and navigation (POS/NAV). Low cost inertial measurement units (IMU) based navigation systems have the disadvantage of accumulating increasingly continuous errors in great extension, leading to poor system performance. Moreover GPS does not work in all environments, or can not provide reliable solutions, under certain circumstances, during some time interval. The integration of both systems can handle and overcome their limitations. There are different solutions to fulfill information during GPS blockage and integrated systems with inertial sensors and GPS are frequently used with stochastic parameter estimation techniques. This work investigates the use of artificial neural network (ANN), to learn and compensate for IMU errors such as to provide better NAV/POS solutions, during the lack of information in GPS outages portion of time. It is proposed to model the input-output ANN signals based on a set of constrained land vehicle navigation equations and the use of an adaptive ANN training Kalman filtering methodology. Numerical simulation results are presented based on urban vehicular positioning application data, acquired from low cost Crossbow CD400-200 inertial measurement unit and an Astech Z12 GPS receiver. Comparison with equally simulated results of a current more standard Kalman filter INS/GPS integration scheme gives a indication of feasibility and potential of the proposed approach.
Palavras-chave: Low cost navigation; ANN; inertial navigation; GPS; IMU; Kalman filter
Código DOI: 10.21528/CBIC2011-14.1
Artigo em pdf: st_14.1.pdf
Arquivo BibTex: st_14.1.bib