Path Recognition And Optimal Control Of Agricultural Automatic Guided Vehicle Based On Vision Navigation

Field intelligent vehicle for agriculture are one way to improve the automation of agriculture, which can be satisfied with the demand of future agriculture with high yield, food safety, high working efficiency, environment protection and precision horticulture. The automated guidance technology is one that must be dissolved as the basic technology. The paper has focused on the calibration algorithms of vision system, algorithms of field road recognition, extracting the navigation parameters, automatic guidance controller.1. The calibration of camera. In order to meet the precision of automated guidance, we use lens radcial distortion model combine with Tasi two-stage method to calibration the camera of vision system. To extract the pixel coordinate of circle target, we can compute camera datas in-and-out. Tested by the already calibrationed parameters, the precision of calibration can granted the guidance demands..2. Image Pre-processing and extracting of crop row. First, we adopt the media-filter to remove the noise of images is better; using the normalized excess-green feature and the fixed threshold to segment images of cucumbers grown in rows in greenhouse. Several mathematical morphlogy methods is used to remove the holes,then extract the boundary of plantsAt last we obtain the crop rows with boundary.3. Extracting the navigation parameters.In order to meet real-time and presicion control,we use Hough transform combining with regions of interest to extracting the navigation parameters. This method has been tested feasible in real fields.The mothod not only quicken the extracting speed but also minish the lateral and heading error.4. Designing controller. AGV kinematics model is founded based on AGV's structure features and relative motion relations with lane. Driving system identify is finished using least square means and the driving system's transfer function is got. State space model of the system is founded. The paper chooses the linear quadratic form optimal controllerdesign scheme in the contemporary control theory by applying the modern optimal control theory. Then obtain the optimal state feedback array K, and finish design of optimal controller of AGV. Simulation results show that the designed controller can realize accurate and stable navigation control and it has the well steady-state characteristic and adaptability to the changes of speed.