Research On Collision Detection Of Sensor-less Manipulator Based On Wavelet Analysis

Man-machine cooperation is the main development direction of industrial robot.The arrival of the Industry 4.0 has further shortened the distance between humans and robots.The safety of the robot can not be ignored,through effective collision detection and correct response is an important guarantee for the safe operation of the robot.From the electromagnetic torque formula,we can get the characteristics of the correlation between the current and the Torque,and the fluctuation of the current itself will correspond to the change of the output torque.The current loop is the fastest part of the robot’s three-ring control system.However,in the actual work of industrial robot,the current signal produced by the motor is accompanied by noise interference,which will affect the accuracy of detection.Therefore,a threshold de-noising method based on wavelet transform is proposed in this paper,which is very suitable for dealing with non-stationary signals.The following work was carried out in this study:1、According to the electrical characteristics and mathematical model of servo motor,the simulation model of servo motor can be established in Simulink,and the relationship between current and output torque can be found.2、Set up the current acquisition experiment platform in order to collect Acquisition Motor Collision Phase Current in real time,and data into Matlab software,in order to analyze and process the signal.3、Get the frequency characteristic of the signal by the method of spectrum analysis,and choose the suitable de-noising method.4、Through the wavelet transform toolbox in Matlab to analyze the collected data, in order to decompose layer number,wavelet base,threshold value,threshold function four aspects in detail to explain how to select the problem,the fixed threshold denoising method and the variable threshold denoising method are compared to get the difference of the two denoising effects.