Tahereh Taleshian - Electrical Engineering Dept., Babol University of Technology, Babol, Iran
Abolfazl Noei Ranjbar - Electrical Engineering Dept., Babol University of Technology, Babol, Iran
Reza Ghaderi - Electrical Engineering Dept., Babol University of Technology, Babol, Iran

In this paper, an integration of Improve Particle Swarm Optimization (IPSO) incombination with Successive Quadratic programming (SQP) so called IPSO-SQPalgorithm is proposed to solve time optimal bang-bang control problems. Theprocedure is found not sensitive to the initial guess of the solution. Due to randomselection in the first stage of the search process, the chance of converging to theglobal optimum is significantly increased, without sticking in a local optimum. Thecombined technique gains both advantages of its original algorithms. The IPSOdirectly minimizes the cost function without the need for gradient-based techniques.The performance of the outcome will be increased when the SQP immediatelyundertakes the optimization task. This is shown via applying those on some othernonlinear systems. Consequently, the proposed algorithm is successfully applied ona time optimal bang-bang control of an autonomous underwater vehicle. A pitchprogrammingtask is also investigated for the autonomous underwater vehicle bydesigning an optimal PID controller

Autonomous Underwater Vehicle, IPSO-SQP Algorithm, Optimal PID Controller, Pitch Programming, Time Optimal Bang-Bang Control