3륜 전방향 무인운반차량용 다입력 다출력 강인서보제어기의 개발

Abstract

Today, automated Guided Vehicles are being used for several factories and modern hospitals. Automated guided vehicle (AGV) applications in the automotive industry include automated raw material delivery, automated work in process movements between manufacturing cells, finished goods transport, and works in dangerous environment, etc. Specially, omnidirectional AGV (OAGV) is one of principal requirement for AGVs designed for works in narrow area and in dangerous environment such as health general-hospital services, works in nuclear power plants, etc. This thesis is about development of a robust servo controller for a MIMO OAGV system for tracking desired trajectories using laser sensor. To do this task, the followings are done. Firstly, a structure of the OAGV used for experiment is described. The OAGV consists of platform, three independent driving omnidirectional wheels which equally are spaced at 1200 from one another, and several interconnected devices. A real OAGV system is developed with several interconnected devices such industrial computer as main controller, monitor, keyboard, mouse, DC servo drivers, DC servo motors, encoders, laser sensor NAV-200, etc. A laser sensor device NAV-200 is used to detect the OAGV posture in indoor environment in real time. Secondly, mathematical modellings of the three-wheeled driving OAGV are proposed. kinematic modelling of the three-wheeled driving OAGV with equal geometric center and gravity center is proposed and dynamic modelling of the three-wheeled driving OAGV with a disturbance vector such as uncertain friction and slip is also proposed. Thirdly, based on these mathematical modellings, a MIMO robust servo controller using a polynomial differential operator(PDO) is proposed. To design the proposed MIMO robust servo controller, the followings are done. Firstly, the posture vector of the OAGV is measured using laser sensor NAV-200. Secondly, an output error vector from an output vector and reference input vector is defined. Thirdly, an extended system and a new input law are obtained by applying the PDO to the output error vector and the system state space equation. Fourthly, a feedback control law is designed by the regulator design method using the pole assignment method. Finally, simulation results are shown to verify the effectiveness of the proposed MIMO robust servo controller. Finally, based on these mathematical modellings, a MIMO robust servo controller using a linear shift invariant differential operator(LSIDO) is proposed. To design the proposed MIMO robust servo controller, the followings are done. Firstly, the posture vector of the OAGV is measured using laser sensor NAV-200. Secondly, an output error vector from an output vector and reference input vector is defined. Thirdly, an extended system and a new input law are obtained by applying the LSIDO to the output error vector and the system state space equation. Fourthly, the proposed MIMO robust servo controller is designed by the regulator design method using the pole assignment method. Fifthly, a servo compensator and a control law of the MIMO OAGV system using inverse LSIDO. Finally, simulation and experimental results are shown that the proposed MIMO robust servo controller has the adequate tracking performance under a step type of external disturbance and the high-order reference signals such as step, ramp parabola. These results are compared with those of the sliding mode control method proposed by N. Hung in 2010. The proposed MIMO robust servo controller demonstrates faster tracking performance than the sliding mode control method.