Study On The Cleaning And Removal Technology For Autonomous Mobile Robot's Vision Based On Coanda-Effect

Due to the electrical property and the dust surface which area generated by dust are many and large, they make the adhesion forces likely to be stronger and stick to the robot camera easily. The dust is difficult to drop out automatically and resulting in the robot vision assignment and monitor environment to be carried out unsuccessfully. Thus, it is an urgent problem to clean the robot camera cleaning device automatically. In this paper, the following study has been done for the dust removal technology of wheeled robot vision system:Firstly, according to the requirements of the camera shape and installation space of the vision system of the wheeled robot, a preliminary program used with air was designed based on Coanda-Effect. High speed air flow is being input from one side of the channel, there is more air filled in the cavity result from Coanda-Effect which also can save air flow.Secondly, GAMBIT was used to establish fluid model. The boundary condition was set .The numerical analysis was done by FLUENT. In the research process, the fluid was divided into two parts which include attached flow field and chamber flow field. In the structure design of chamber flow field, the symmetry axis of the inlet then offset along the symmetry axis of the inlet were studied. The flow velocity of the outlet slit monitoring point and the linearity index is utilized to describe the degree of flow velocity. In the design of attached flow field, aiming to the changes of the radius of the Coanda outlet surface, the flow amplification and inlet-outlet velocity were studied to obtain result from the modification.The difference of inlet velocity also was analyzed for the flow amplification and inlet-outlet velocity. At last, the optimum size of inlet Coanda radius was studied, too.Thirdly, the structure size was optimized based on CAO software ISIGHT integrated with GAMBIT and FLUENT. The structure dimensions of the attached flow field and chamber flow field were taken as design values and the linearity of the flow velocity of the outlet slit monitoring point and the average flow velocity were regarded as the objective function of the chamber flow field. The mirror mean velocity and the flow magnification were treated as the objective function of the attached flow field. The above, their structure dimensions were optimized. The effect of objective function namely contribution rate of every value was analyzed based on the DOE method. At last, the MIGA algorithm was used for structure dimension optimization.Finally, the prototype of cleaning cover was built based on the above theory. The experimental apparatus and experimental arrangement were introduced at the beginning of the experiment. Following that regards, the flow amplification experiment and flow path experiment were validated by the experiment. In conclusion, the dust elimination to the camera device of wheel robot vision system is verified by experiment. The upper computer was compiled by VB for the visual feedback system. The different radius of the dust were used to validate and analyzed the cleaning effect under the condition of five different air flow velocity.