Estimating Immeasurable Variables of Quadrotor Using PSO Based Virtual Sensing System

Abstract

In recent years, the implementation of intelligent control systems in many control theory lead to the development of virtual sensing technologies which can predict most of the systems states accurately. Virtual sensing system or virtual sensor which consists of a Diagonal Recurrent Neural Network (DRNN) coupled with Extended Kalman Filter (EKF) will estimate the immeasurable system variables using easy to measure known variables. In this paper, virtual sensor scheme which is optimized by Particle Swarm Optimization (PSO) algorithm is introduced. The new scheme will be used to estimate the position parameters of a quadrator, which originally has all 12 state variables to be determined. Simulation studies reveals that the designed virtual sensor scheme is able to predict the positions state of the quadrotor accurately.