Research On Motion Control Of Manipulator Of Hydraulic Excavator

Research on intelligent and robotic excavator has become a focus in a world, and this type of excavator becoming more and more important in application. Under this background and combining the national 863 project, the aim of thesis is researching on accurately motion controlling of robotic excavator's manipulator. Beginning from the kinematic, dynamic and electrohydraulic system models of excavator's manipulator, the theoretic researches of identification, prediction and control are developed based on the CARIMA model. Then, the result of theoretic research is applied to robotic excavator's manipulator, the main research work as follows:The development of research on motion control of excavator's manipulator is surveyed. The function and general scheme of excavator are given according to the requirement of research, and the robotic excavator is produced.Full kinematic and dynamic models of the excavator arm are derived, Matlab programs and simulation of models are offered. The three dimensions solid model for excavator arm is derived exactly by combining experiment with theoretical research. Based on the solid model, kinematic and dynamic simulations are also presented in ADAMAS. Both of the simulation results are consistent.On the request of motion control, the different trajectories plan of excavator's manipulator are given, which are level, inclining, vertical and are lines. The simulation of path plan is also given.The principle for LUDV system of robotic excavator is analyzed, and then the flow equations are developed. Starting with the basic equations of hydraulic systems and according to experiment and practice, a simplified model of eleetrohydraulie system for excavator's manipulator is obtained. The estimation methods and formulas for key parameters in model are given. Then the four model forms of input-output continuous-time transfer function, continuous-time state space expression, discrete-time state space expression and discrete-time input-output model are deduced. A number of control methods are developed step by step, including PID control of 2 DOF (degree of freedom), PID control of 3 DOF, and model reference adaptive control for the manipulator, and the results of experiments for them are given. It is shown that the accuracy of PID control can meet the requirement when the speed of bucket is less than 136mm/s. the latter can improve the speed of bucket, but the accuracy of it is low when the reference trajectory is complex.Because of the disadvantages of the above methods, the Self-Tuning Regulator base on the online identification for the parameters is chosen for the motion control of excavator's manipulator. Firstly, the least square estimation fitting for excavator's manipulator is developed, which is combining the forgetting factor algorithm with extended matrix. The simulation in Matlab is presented, and results show that this least square estimation is valid.Then, the estimation of CARIMA for the electrohydraulic proportional system of excavator's manipulator is designed; the j step predictive control for the model is developed. The design step for adaptive predictive control is obtained. The simulation research on nonlinear system and the electrohydraulic proportional system of excavator's manipulator is implemented respectively. It is shown that the method is feasible. Especially, the adaptive predictive control with variable weight parameter for the electrohydraulic proportional system of excavator's manipulator is deduced during the simulation. In the following research, it is this method that can accurately control the motion of manipulator.At last, the experiment for the adaptive predictive control is demonstrated in order to verify the above theoretic research. Using the adaptive predictive control in paper, it is shown whatever for the single trajectory or multiple trajectories the accurate tracking is reached. All of the errors for trajectories tracking are less than 100mm; moreover, the method in paper is applied to the practical project. The result is good too. It is undoubtedly show that the adaptive predictive control designed in the paper is feasible. In a word, the paper provides a complete theorem and method for the motion control of excavator's manipulator, and the research is valuable and can be popularized.