Research On Robot Force Control Method Used On Grinding Process

Teaching and playback is the main mode of grinding robot’s application field. The grinding trajectory cannot adjust according to the errors, like form and position error of workpiece, clamping error and robot positioning error, in the process of operation, which leads to uneven quality of the workpiece, but also limits the scope of application of grinding robot. In order to achieve the high quality and high consistency of robot grinding, the process of contact force control is essential. Using force sensor and other equipments to obtain contact force information is needed, and an appropriate force control algorithm to adjust the machining trajectory needs to be designed. A uniform material removal rate could be obtained by keeping the grinding pressure constant, so that it can enhance the degree of automation of the process robot. Aiming at the problem of robot force control, this thesis makes a series of studies, including the force signal measurement and processing method, constant force tracking control, force controlled grinding, the realization of real-time force control system.Firstly, the method of measurement and signal processing are studied. The formula of force coordinate transformation based on wrist force sensor is derived. The decoupling matrix of Maxwell force sensor is calibrated. Considering the contradiction between sensitivity and stability of first-order inertial filter, the adaptive inertial filter is used.Secondly, the unknown surface contour tracking with constant force control method is studied. A double loop control method for contour tracking is proposed. Double-loop in this method including movement angle adjuster based on real-time feedback, and the nonlinear PID normal contact force controller. Simulation and experiment results have proved that the double loop control method is feasible, and has a higher control precision of normal contact force.Thirdly, in order to avoid the noise amplification effect in the differential link of traditional PID controller, an improved PID control method is designed. To achieve the desired dynamic response, the control parameters are determined by the closed loop dominant pole method. The optimization rules are established based on the matrix encoding genetic algorithm. The fuzzy rule matrix is optimized by the rules. And the optimized Fuzzy-PID controller is used to control the grinding force. Simulation and experiments show that the optimized Fuzzy-PID controller has high dynamic response quality and strong adaptability.Finally, In order to increase the force feedback control of the position controlled robot, the RB08 industrial robot force control system based on ARM and SRI force sensor is designed. The experiment results show that force controller based on ARM can realize the real-time force feedback control. In addition, without changing the original position control model of industrial robot, this method has a large application reference value.