Experiments of real time stochastic controller using approximate PDF

In the past decades, various stochastic control techniques has been introduced to guarantee the stability and to reduce the perturbation of system under random disturbances. However these techniques have imperfections of not using its stochastic information of the disturbances fully in designing control system. Due to the reason, ´Heo-stochastic controller´ is developed very recently and is proven to show better performance than the conventional stochastic controllers such as LQR, LQG do. Nevertheless ´Heo-Stochastic Controller´ has some realistic limitations to realize it. System equation in time domain is transferred to dynamic moment equation in stochastic domain and ´Heo-stochastic controller´ is designed. Two problems have been confronted with in the process of design ´Heo-stochastic controller´. One is problem of real time implementation of the controller and the other one is problem of order increasing in stochastic system equation. In order to complement both disadvantages, approximate probability density function is adopted in the design of ´Heo-stochastic controller´. Wind tunnel experiment is conducted, where very low subsonic turbulent flow is generated to imply on model helicopter system in side way. Stochastic pitch control of vertical propeller system is successfully demonstrated by using proposed concept. Modified band limited white noise approach is also utilized in the course of both simulation and experimental studies.