Navigational Strategies for Control of Underwater Robot using AI based Algorithms

Kundu, Shubhasri (2015) Navigational Strategies for Control of Underwater Robot using AI based Algorithms. PhD thesis.



Autonomous underwater robots have become indispensable marine tools to perform various tedious and risky oceanic tasks of military, scientific, civil as well as commercial purposes. To execute hazardous naval tasks successfully, underwater robot needs an intelligent controller to manoeuver from one point to another within unknown or partially known three-dimensional environment. This dissertation has proposed and implemented various AI based control strategies for underwater robot navigation. Adaptive versions of neuro-fuzzy network and several stochastic evolutionary algorithms have been employed here to avoid obstacles or to escape from dead end situations while tracing near optimal path from initial point to destination of an impulsive underwater scenario. A proper balance between path optimization and collision avoidance has been considered as major aspects for evaluating performances of proposed navigational strategies of underwater robot. Online sensory information about position and orientation of both target and nearest obstacles with respect to the robot’s current position have been considered as inputs for path planners. To validate the feasibility of proposed control algorithms, numerous simulations have been executed within MATLAB based simulation environment where obstacles of different shapes and sizes are distributed in a chaotic manner. Simulation results have been verified by performing real time experiments of robot in underwater environment. Comparisons with other available underwater navigation approaches have also been accomplished for authentication purpose. Extensive simulation and experimental studies have ensured the obstacle avoidance and path optimization abilities of proposed AI based navigational strategies during motion of underwater robot. Moreover, a comparative study has been performed on navigational performances of proposed path planning approaches regarding path length and travel time to find out most efficient technique for navigation within an impulsive underwater environment.

Item Type:Thesis (PhD)
Uncontrolled Keywords:Underwater robot, Three-dimensional Navigation, ANFIS, Shuffled Frog Leaping Algorithm, Differential Evolution, Harmony Search
Subjects:Engineering and Technology > Mechanical Engineering > Robotics
Divisions: Engineering and Technology > Department of Mechanical Engineering
ID Code:7389
Deposited By:Mr. Sanat Kumar Behera
Deposited On:20 May 2016 18:50
Last Modified:20 May 2016 18:50
Supervisor(s):Parhi, D R

Repository Staff Only: item control page