| In order to improve the automation and intellectualization level of brush cutter,realize the automation work in the forest farm,reduce the working intensity of the driver of brush cutter,reduce the repetitive work in the operation of brush cutter,and effectively solve the problem of damage to young trees caused by human errors in judgment or operation.Based on the analysis of the working environment of Eucalyptus plantation and the operation situation of the clearing machine,this paper puts forward a method of obtaining the two-dimensional characteristic information of the stand by the laser radar recognition technology,and establishes the recognition system of the stand information.Through the kinematics analysis and modeling of brush cutter,the aim is to analyze the effective control strategy of brush cutter.Combining with the effective and stable tree information provided by the recognition system,the motion control system of brush cutter is constructed.The main contents of this paper are as follows:1.Through the analysis of the existing cutting machine prototype in the laboratory and the previous research on the two-dimensional laser radar measurement,the existing two-dimensional laser radar is selected to build the object feature extraction system.2.For the two-dimensional laser radar point cloud data acquired by forest farms,various filtering schemes are used to preprocess the data.Combining with the real data,the appropriate point cloud data processing scheme is selected.In order to effectively distinguish the point cloud data of laser radar,several segmentation schemes are analyzed,and the segmentation effect of point cloud data is compared.The method based on distance threshold is adopted.For the segmented target,in order to effectively distinguish the stand and shrub weeds,the laser radar is installed at a certain height to exclude the influence of low shrubs.For a certain height target,the method of circular fitting is proposed.The diameter of standing trees is generally thicker than that of shrubs.The location information extracted by tree recognition provides data support for the movement control of the cutting machine.3.Aiming at the long row spacing of Eucalyptus plantation generally according to certain planting rules,this paper also analyzed the lateral pendulum brush cutting method which is widely used in the brush cutting of forest farms,and proposed that the brush cutting brush cutting method of the cutting and brush cutter can be intelligently controlled between the rows of Eucalyptus.In order to realize intelligent control,firstly,the D-H model kinematics model of the cutter is established,and the relationship between the cutter's actuator and the hydraulic cylinder length is obtained,which provides theoretical basis for the accurate control of brush cutter.Aiming at the chassis control problem in the process of brush cutting with horizontal pendulum brush cutting,a two-dimensional laser is proposed to acquire the position information of trees between rows,which provides parameters for chassis direction control.Aiming at the problem of trajectory planning in the process of rotary brush cutting,a method of rotary angle control of clearing machine was proposed.The distance between rows of trees was matched with the operating radius of brush cutter,and the deflection angle of rotary mechanism was obtained as the angle control range.In order to solve the problem of tree collision in cutting head trajectory planning,a control scheme based on the relationship between distance and azimuth of standing trees was proposed.The validity of laser point cloud data processing method is verified by experiments.The proposed segmentation algorithm and fitting algorithm can effectively deal with the extraction and recognition of targets.The effectiveness of the trajectory control motion model and obstacle avoidance method of the cutting head is verified by Matlab simulation experiment.Combined Matlab and V-rep software to design the simulation control system of brush cutter. |
PDF Full Text Request