Platform and steady Kalman state observer design for Intelligent Vehicle based on visual guidance

State observer design is one of key technologies in research field of intelligent vehicle. Experiment platform, visual guidance intelligent vehicle JLUIV-5, is established by Jilin University Intelligent Vehicle Group firstly. The system structure and assistant navigation control system, and different image identify algorithms to recognize preview path and stops for variable illuminations are introduced. The dynamic response equation of steering control system was got by system identification experiment. By combined with the preview kinematics model, and two-degree steering dynamic model of vehicle, the steering Kalman filter mathematics model based on preview kinematics for intelligent vehicle was obtained. And observer is designed by applying steady Kalman filter theory. The simulation and experiment results, carry out in Jilin University Nanling Campus and Culture Center of Jilin Province, show that the image identify algorithms, and steady Kalman state observer designed by the proposed method has good adaptability for time-varying and parameters uncertain, it can satisfy intelligent vehicle trace the path reliably during outdoor experiment.