Research On Mechanism And Control System Of Manipulator Of Logging Harvester

Logging harvester with full-function of felling, delimbing, bucking and so on is the symbol of modernization of forestry machinery. But no longing harvester has been used in China. Under this background, relying on the "11th five-year plan" national science and technology support plan of funded project (project Numbers 2006BAD11A5) "The key technical of Multi-function forest cut and rear work equipment research and development", this research of automatic control system is based on the CFJ-30 logging harvester developed by the group.The main purpose of this study is to realize the automatic control of positioning, felling delimbing, bucking for manipulator. First, the felling head trajectory is determined.by the analyse of kinematics and dynamics of the manipulator, and together with the visual information from 3d laser scanner and the angular displacement information obtained by angle sensor. Then, by analysing the electrohydraulic proportional system, the relation between the current value of electrohydraulic proportional valve and the motion trajectory of the felling head can be obtained, meanwhile the current is discrcted according to the same time intervals. Finally, the modified discrete points by inputed into the neural network program were readed by a self-manufactured singlechip controller. Thus, the automation positioning of the felling head can be realiazed by the control of electrohydraulic proportional valve from the readed discrete points. In order to further realize the automation control of felling delimbing and bucking, directional valves must to be controlled. In this paper, voltage value controlling reversing valve were discreted according to same time interval and meanwhile the discrete points were readed by the singlechip controller. Thus, the automation control of felling delimbing, bucking for manipulator can be realized. The specific works are as follows:The present and the development tendency of domestic and foreign logging harvester and its manipulateor were described. According to Chinese forestry machinery research situation, automation research direction of the logging harvester manipulator were determined to realize the ntelligent control.The structure and function parameters of the logging harvester were introduced. The structure of the manipulator and structure and working principle were analyzed in order to to lay the foundation for latter research.Using D-H method to establish the rod coordinate system of the mechanical arm, according to which, the transformational matrix between the coordinate system of end actuator of the mechanical arm and the base coordinate can be obtained. The analytical form of each joint of manipulator can be obtained by inverse transformation method and the conversion between drive space and joint space can be got through the sport's mechanism kinematics. Using Lagrange method to establish the dynamics equation and solve it. Planing the motion trajectory of the manipulator and simulating manipulator by the software LSM Virtual Lab. in order to provide the theory basis.for control system design.Analyszing the hydraulic system of the logging harvester, building electrohydraulic proportional hydraulic system model, estimating load of master cylinder and minor cylinder, building the state space model and discrete model of the harvester electromechanical liquid proportion, the relation between the current of the electrohydraulic proportional valve and trajectory of the felling head can be established.To realize the intelligent control, the visual system of the harvester was built. The point cloud data obtained by 3d laser scanner were simplified, partitioned and conducted coordinate transformation. And then the position, entroid radius and the height of the target tree can be calculated and provides data for the neural network control system.By combing neural network theory with modern control theory of the robot, a new method of control the harvester mechanical arm was brought up based on fuzzy adaptive kalman filter radial basis function neural network. Because the current value got by the path planning and electrohydraulic proportional system have more uncertainty, the obtained result must be modified by input into the neural network in order to reduce error and to increase the level of the intelligent control. The accurate degree of positioning will.be greatly improved by using the singlechip controller to read the modified discrete values.In order to achieve the automatic control of the harvester better, a singlechip controller was developed, and its control objects were respectively electrohydraulic proportional valve and reversing valve.Using discrete points output by the neural network can control the electrohydraulic proportional valve movement, thus the felling head can move according to precalculated trajectory. Using the voltage discrete points read by the singlechip can control the movement of the reversing valve, thus the automatic control of felling delimbing, bucking can be realized. In this paper, according to the characteristics and function of the control objects, the singlechip was developed and when the discrete points mentioned above were input into the singlechip, the automation positioning, felling delimbing and bucking can be realized.Finally, the instruments and equipment the automation control needed was introduced and the test was conducted by utilizing the automatic control systcm in order to prove the effectiveness of this study.