Nonlinear Model for Autonomous Underwater Vehicle and Identification Based on Least Square Algorithm

A nonlinear model for Autonomous Underwater Vehicles is proposed. In order to describe a more precise dynamic behavior, the nonlinear model for both Lateral and Longitudinal subsystems is derived based on all applied forces and moments. The proposed model can be explained as an extended linear model for AUV in depth and azimuth motions where some nonlinearities are taken into account. Due to some practical issues as well as the form of proposed model, the identification problem based on Least Square method is formulated to achieve the system parameters. By considering unstable dynamic of system, the open loop system cannot be excited. In this case, the PID regulators with simple tuning parameters are proposed in both Lateral and Longitudinal subsystems and the identification problem by utilizing sinusoidal inputs is followed within a feedback loop. Based on measurable variables i.e. linear moments, angular velocities and Euler angles, and utilizing some dynamic filters, the Least Square method is then applied to estimate the model parameters. The effectiveness of proposed nonlinear model as well as the parameter identification approach are finally demonstrated through some numerical simulations.