Multifeedback-layer neural network controller design using particle swarm optimization algorithm for hard disk drive control

In this paper, the weights of the Multifeedback-Layer Neural Network (MFLNN) which has recently proposed are trained by the Particle Swarm Optimization (PSO) algorithm. To improve training capability of the PSO, it is enhanced by some modifications. This method (MFLNN-PSO) is applied to two different problems to prove accomplishment of the method. Then, the closed loop identification of the reader head position of a disk drive system is proposed by using the MFLNN-PSO algorithm. Finally, a new type of neuro controller is put forward by using the MFLNN-PSO. Initially, this neuro controller is applied to two different kinds of dynamic systems. Later, it is applied to a hard disk drive system as a real physical example. Simulation results show that the MFLNN-PSO controller is effective and efficient on the control of dynamic systems and hard disk drive system.