In this paper, an integration of Improve Particle Swarm Optimization (IPSO) in combination with Successive Quadratic programming (SQP) so called IPSO-SQP algorithm is proposed to solve time optimal bang-bang control problems. The procedure is found not sensitive to the More
In this paper, an integration of Improve Particle Swarm Optimization (IPSO) in combination with Successive Quadratic programming (SQP) so called IPSO-SQP algorithm is proposed to solve time optimal bang-bang control problems. The procedure is found not sensitive to the initial guess of the solution. Due to random selection in the first stage of the search process, the chance of converging to the global optimum is significantly increased, without sticking in a local optimum. The combined technique gains both advantages of its original algorithms. The IPSO directly minimizes the cost function without the need for gradient-based techniques. The performance of the outcome will be increased when the SQP immediately undertakes the optimization task. This is shown via applying those on some other nonlinear systems. Consequently, the proposed algorithm is successfully applied on a time optimal bang-bang control of an autonomous underwater vehicle. A pitchprogramming task is also investigated for the autonomous underwater vehicle by designing an optimal PID controller .
Manuscript profile
