The Design Of AUV PID Controller Based On MOOS And ELM

Nowadays, with the autonomous underwater vehicle (AUV) being widely used in the seafloor resource exploration and military science category, the intelligence of AUV has become an important aspect of people’s attention and research. Decision-making and Control is one of AUV intelligent reflection, it’s the key to ensure the successful completion.This thesis designs a PID controller based on ELM which can do a good job about controlling dynamic system of AUV.This thesis firstly introduces the applications of AUV at home and abroad,and the architecture of a torpedo-shaped AUV as well,then,focuses on responding behavior-based Decision-making Control system.In the thesis, the Decision-making Control system of AUV is divided into two kinds:rational behavior subsystem and perceptual behavior subsystem according to behaviors. The rational behavior owns high intelligence and complex algorithm, the perceptual behavior is the reaction of AUV to outside world,whose algorithm is simple, real-time is high, set of the perceptual behavior constitutes a the perceptual behavior subsystem, two subsystems work together and complement each other, to ensure AUV completing the required tasks excellently.Secondly, this thesis introduces the MOOS (The Mission Oriented Operating Suite) architecture, which has three prominent characteristics:modularization, distributed, star topology. The torpedo-shaped AUV software systems research of this thesis isdesigned based on MOOS architecture, according to the advantage of modular design, good code reusability, high scalability characteristics, the vehicle carry a number of functional modules based on requirement, focus on the development case aboutDoppler Velocity Log (DVL) sensor module and PID control module based on ELM.Finally, Extreme Learning Machine (ELM) is a new Single-hidden Layer Feed forward Neural Network, with high learning rate, good training and other advantages, PID controller based on ELM in this thesis is able to find the optimal control factor to achieve controlling the AUV more precisely. After that, using the AUV heading as control object, the thesis get the simulation about output signal of the PID controller based on ELM, Through analysis discovery of simulation results and research, it indicates PID controller design based on ELM can control AUV more accurately travel according to the planned route, then successfully reach the destination.