Reach on Control System of Vision-based Accelerate Loading Testing Vehicle
|School||Harbin Institute of Technology|
|Course||Mechanical Manufacturing and Automation|
|Keywords||accelerated loading testing vehicle vision navigation wirelesscommunication robust control|
Accelerated loading testing vehicle is a mechanical vehicle used for testing theperformance of pavement under the facility, it can simulate the rut formation resultof the testing pavement under different acceleration and load. The traditionalaccelerated loading testing vehicle uses the manned way to control the vehicle, tomake the facility more automatic, this paper put forward a vehicle control methodby using the technology of vision navigation in the control system of acceleratedloading testing vehicle.According to the analyzing of the experiment environment in acceleratedloading test, a architecture of total distributed based on Multi-Agents was proposedfor the vehicle control system; on the basis of this architecture, a general frameworkof control system for the vehicle was established and the working principal of thecontrol system was analyzed; contraposing the closed environment of the test, thezigbee model was used to designed the wireless communication network and thesoftware; to complete the task of image processing for vehicle navigation, theDM6467dual-core processor was put to use as the image processing platform; adeveloping environment for the platform was put up and a series of tasks includedeveloping a multi-thread image capture and manipulation software, assembling thealgorithm engine in ARM and the algorithm server in DSP has been done; aimproved method of navigation parameters extraction was put forward based on thecontrast analysis of sobel operator and laplacian operator and binary converting ofthe road image by iteration algorithm; according to the analysis of the vehicledynamic process, the dynamic model of the vehicle was set up to establish theclosed-loop feedback robust control framework, the method of μ-synthesis wasadopted to design the robust controller k of the vehicle and the running process ofthe vehicle was simulated by the matlab toolbox under the robust controller k; toverify the tracking effect of the navigation algorithm, a small vision navigationvehicle platform was constructed.The result of the simulation and experiment indicated that the control systemdesigned for the accelerated loading testing vehicle has a good tracking effect, fastresponse and high robustness; the wireless communication network can transfer thedata accurately in realtime; the improved method of the navigation parametersextraction can track the navigation path exactly.