Reseach On Compliance Force Control For Robot End Pneumatic Actuators

The automation of the industrial production line is getting higher and higher,especially in the surface treatment of the workpiece and the different types of machining,such as polishing and cutting.In such a process,industrial robots are widely used instead of manual operations nowadays,and the machining quality of the workpiece is directly affected by the relative force of the contact surface.Compliance control can help the mechanical structure complete some compliance actions and make the whole system intelligent.The general classical control algorithm can achieve satisfactory control result in a fixed environment.However,it make the general control algorithm difficult to ensure the robustness of the system because of the complexity of the environment and the system itself is nonlinear and time-varying in practical applications.To solve this problem,the internal model control is used to ensure that the system is robust to model errors.At the same time,the control requirements in complex environment of the pneumatic system are realized through the improved internal model control structure and filters,as well as the online identification algorithm based on the fuzzy dynamic model.Firstly,the cylinder and the high-speed solenoid valve the main components of pneumatic system are analyzed and modeled.The control on flux compensation of the solenoid valve is carried out,which reduces the nonlinearity and modeling inaccuracy of the whole system.Then the approximate linear transfer function of the system is obtained by reasonable selection of parameters based on previous experimental experience,and a preliminary analysis is made.Secondly,the internal model control method is applied to the pneumatic system.The advantages and disadvantages on tracking and anti-disturb performance of the classical internal model control are pointed out.At the same time,the condition of desaturation and working without tracking error are theoretically obtained through analysis of tracking performance under saturation.Next to it,the classical internal model filters are improved,and its improved tracking and anti-disturb performance are confirmed.Then the anti-windup controller with model error which improves the anti-windup performance of thepneumatic system is designed by using the condition technology under the anti-windup internal model structure.Further more,the fuzzy adaptive internal model control based on the online identification in the continuous domain is proposed.The fuzzy dynamic system adopted to approximate the actual nonlinear time-varying pneumatic system overcomes the shortcomings of the classical internal model control,and the stability of the algorithm is proved strictly.Finally,the all algorithms are verified by the active control flange experimental platform.The varying input and anti-disturb experiment of the classical and the improved internal model control experiment are designed.The effectiveness and stability of these algorithms are verified through comparing the performance of the two.What's more,it is proved that the improved internal model controller is more excellent than the classical.