Surge Motion Parameter Identification for the NPS Phoenix AUV

Recent interest in underwater vehicle maneuvering and control in shallow water has generated a need for greater understanding of vehicle dynamics in this region. Specifically, improved vehicle models and control methodologies are required for acceptable control performance in the presence of wave induced disturbances. In this paper a method for identifying the decoupled surge motion dynamic parameters of a small Autonomous Underwater Vehicle (AUV) is given. The identification is based on experimental data obtained from in-water tests of the vehicle along with post processing of the data using Kalman filtering techniques. Various lumped parameter system models of the surge dynamics are proposed and comparisons are made between them with the actual motion measurements. Identification of the square law velocity, square law propeller speed coefficients, and any first order force lag time constants is performed, as well as selection of the most appropriate model design for the experimental conditions. Since a continuously changing surge motion is required for adequate parameter identification, the vehicle is placed in a test tank and commanded to oscillate in the longitudinal direction using the stern propulsors. A series of experiments are performed with different sets of commanded oscillation amplitudes and frequencies of motion. Surge position and rate feedback for the vehicle controller is provided by an onboard high frequency sonar aimed perpendicular to one of the tank walls. A by-product of the in-water experiments is the observation of the vehicle surge performance in wave like conditions. In shallow water, the vehicle surge motion is particularly pronounced and the design of the experiment allows for the controller performance bandwidth to be studied for the specific propulsor installed. With this information and knowledge of the ocean wave conditions in certain areas, the vehicle control performance may be predicted with some degree of accuracy.