Controller design for a tri-turbofan airship system based on particle swarm optimization algorithm & neural network technology.

The goal of this thesis is to design controller for better control of the tri-turbofan airship system. Instead of using the traditional method, this research uses the particle swarm optimization algorithm to randomly generate and finally find out the right solution for the controller. This research using designed linearization model, nonlinear model and also the Neural Network model. The idea using Neural network here is trying to update the model as close as to the real airship system. We apply the controller generated by particle swarm optimization algorithm on the Neural Network nonlinear model, linear model and also on the neural network model. By comparing the input and output of the airship system, the trained neural network model is really close to the real tri-turbofan airship model and the airship can be hold as the requirement.