Research of underwater navigation on a ROV with structure detection and decontamination

For the development of a novel underwater remotely operated vehicle (ROV) oriented to detection and clean-up, a set of inertial navigation system based on MEMS devices including gyroscope, accelerometer, magnetic compass, depth sensor and microprocessor is designed. In order to suppress the gyro drifts, the complementary filter and quaternion algorithm integrating gyroscope are adapted, as well as the Unscented Kalman Filter is designed to improve the navigation accuracy. In addition, gradient descent algorithm and its role in compensating the quaternions are researched and analyzed carefully. Experimental and simulation results show that the methods can reach reliable, high precise navigation effects.