Adaptive simplified fuzzy logic controller for depth control of underwater remotely operated vehicle

A Remotely Operated Vehicle (ROV) is one class of the unmanned underwater vehicles that is tethered, unoccupied, highly manoeuvrable, and operated by a person on a platform on water surface. For depth control of ROV, an occurrence of overshoot in the system response is highly dangerous. Clearly an overshoot in the ROV vertical trajectory may cause damages to both the ROV and the inspected structure. Maintaining the position of a small scale ROV within its working area is difficult even for experienced ROV pilots, especially in the presence of underwater currents and waves. This project, focuses on controlling the ROV vertical trajectory as the ROV tries to remain stationary on the desired depth and having its overshoot, rise time and settling time minimized. This project begins with a mathematical and empirical modelling to capture the dynamics of a newly fabricated ROV, followed by an intelligent controller design for depth control of ROV based on the Single Input Fuzzy Logic Controller (SIFLC). Factors affecting the SIFLC were investigated including changing the number of rules, using a linear equation instead of a lookup table and adding a reference model. The parameters of the SIFLC were tuned by an improved Particle Swarm Optimization (PSO) algorithm. A novel adaptive technique called the Adaptive Single Input Fuzzy Logic Controller (ASIFLC) was introduced that has the ability to adapt its parameters depending on the depth set point used. The algorithm was verified in MATLAB® Simulink platform. Then, verified algorithms were tested on an actual prototype ROV in a water tank. Results show it was found that the technique can effectively control the depth of ROV with no overshoot and having its settling time minimized. Since the algorithm can be represented using simple mathematical equations, it can easily be realized using low cost microcontrollers.