Predictive control strategy for unmanned vehicle navigation

Due to their increasing usage and importance in various applications research in utilizing microprocessors capabilities and control techniques in improving the performance of automatic guided vehicle AGV has taken place. Various research has been conducted to incorporate a vision capability to such vehicles to capture the surrounding environment of the vehicle such as road profile curvatures to be followed or obstacles to be avoided This capability has been paralleled with developing "intelligent" control techniques which use vision to enable the AGV to give the desired performance. Although in these controllers emulation of the human driver reactions to the environment surrounding the vehicle was attempted, the human natural predictive feature was not considered comprehensively. A human driver, through his vision of approaching road curvatures, and considering the dynamics and limitation of the vehicle, would predict the next locations of the vehicle, accordingly decisions concerning the forward speed and the steering mechanism strategies are decided to set these predictions to their desired path profile. Although the performance of the AGV depends on many factors such as the accuracy of the vision and sensing system in computing future road profiles, the concern of this work is to develop a predictive controller which emulates human like driving reactions in the face of road curvatures