Research On Loading Control Of Pneumatic 3-UPU Parallel Robot

The innovative development of robotics is an important symbol of the progress of national high-tech technology and the improvement of industrial automation level.The 3-UPU parallel robot has the advantages of simple design structure,low cost,good stiffness and load-bearing capacity,accurate control precision,rapid action,etc.In this paper,a pneumatic 3-UPU parallel robot is designed.It has the advantages of convenient maintenance,environmental protection,no pollution and high efficiency.It has high application value and prospect.Firstly,the 3D force sensor,data acquisition card,flow proportional valve and cylinder required by the load control system of pneumatic3-UPU parallel robot are selected.According to the functional requirements,the mechanical structure including universal hinge,moving platform,static platform and related connectors is designed,and the experimental platform of 3-UPU parallel robot loading control system prototype is built.The mechanical properties of the 3-UPU parallel robot are analyzed,and the forward and inverse solutions of the loading force are analyzed.Secondly,through deducing and arranging the gas flow equation of the valve port of the flow proportional valve,the gas mass flow change equation of the cylinder and the force balance equation of the piston inside the cylinder,the third-order mathematical model of the flow proportional valve-cylinder is built,which lays the foundation for the further realization of the loading control of the robot.Thirdly,because of the strong non-linearity of pneumatic servo system,it is difficult to obtain good control effect by using incremental PID algorithm for loading control.With the help of the strong nonlinear processing ability and learning ability of neural network,the single-neuron PID adaptive loading controller and the single-neuron PID model reference adaptive loading controller are designed,and the loading control simulation is carried out,and the comparison with the incremental PID loading controller is verified.Simulation results show that the control effect of the two loading controllers is better than that of the incremental PID loading controller.Finally,through constructing the pneumatic load type 3-UPU parallel robot prototype experiment platform control system,The traditional PID loading controller is compared with the single neuron PID adaptive loading controller,the experimental results show that,the single neuron PID adaptive controller in the step response experiments is faster,the steady-state error is reduced by 1.67 N.For sinusoidal response experiments,amplitude attenuation is reduced by 5% and phase lag is educed by 3.5%.