Adaptive Complementary Filter Based on Least Square to Attitude Estimation of Underwater Vehicle

Accurate Attitude estimation/ localization of robots for research, industrial, military, etc., applications has always been a challenge for designers. The purpose of this research is to present an attitude estimation method based on simple principles of low-pass and high-pass filter in the Complementary Filter with Adaptable gains of the least squares method for combining inexpensive sensor information for navigation of an underwater robot with dolphin motion scenario. The proposed adaptive algorithm; automatically calculates the complementary filter gains according to the measurement error and eliminate the need for manual adjustment or the use of fixed gains according to the device dynamics. Proper  performance of the proposed adaptive complementary filter for estimating the attitude of the underwater robot in field tests; confirms the advantage of this method for compensating for noise and time-varying bias inexpensive sensor.