Noise control in magnetic field via ‘Stochastic System Identification Method’

Study on the reduction of inherent noise in magnetic field has been conducted. Distributed model such as magnetic field is well known for its inadequacy to control in conventional ways. 'Stochastic system identification method' from moment equations via Fokker-Plank-Kolmogorov equation has been developed in inverse manner to figure out unknown system parameters of discrete system under random disturbances. The newly developed concept of 'stochastic system identification method' is implemented on magnetic system, which enables to construct virtual discrete operating system in forms of 2
^nd
order using stationary stochastic system output only. Based on the virtually assumed dicsrete model, system parameters are extracted. In order to reduce the noise, ' Hoe-stochastic controller' is designed for the virtually generated discrete model of magnetic field. Also in order to perform system control in real time frame, approximate probability density function is adopted to design the 'Heo-stochastic controller' and implemented on the system. Feasibility of the idea to control stochastic distributed system via 'stochastic system identification method' and 'Heo-stochastic controller' is confirmed in terms of simulation and experiment as well.