Analysis Of Sampling Delay And Research On Control Method Of Visual Servo Control System

The program of Chinese manufacture promotes the further development of the robotics industry.Since vision sensors are an important part of industrial manufacturing,it is necessary to study the effects of vision sensors on the performance of manipulator control systems.This article starts with the sampling rate of the vision sensor,combined with the discrete visual servo system model,analyzes the performance of the system under different sampling delay,and proposes algorithms with visual field constraints and delay compensation to improve the system performance.First,the basic theory of the manipulator control system is given,including the definition of the manipulator coordinates,the construction of manipulator's D-H parameters model,the position and attitude transformation of the point in space,and the relationship of velocity conversion.The imaging principle of the vision sensor is analyzed,then the interaction matrix used to describe the relationship between the camera speed and the image feature speed is derived and calculated.Further,make an introduction to the visual servo motion control of the manipulator.Secondly,a discrete model of the manipulator visual servo control system is established.Considering the motion performance of the manipulator under non-ideal conditions,a joint incremental model is given.Due to the difference between the camera sampling rate and the manipulator control rate,the delay of image feature is analyzed,and the visual model with delay information is obtained.Then construct a state space model to describe the system,analyze system stability and performance,and determine the stability thresholds corresponding to different sampling rates with the root locus method.Then,based on the different servo results brought by different sampling delay,considering the controller design with field of view constraints,calculate the weight value of the feature at each moment from the view of image feature,define the weight error,then get the input of system to achieve the constraint of visual field.The actual usage image feature is compensated and estimated using the time series and kalman filter base on the image information at the current time.The time delay compensation is realized,and the selection of the stable gain threshold in the designed control algorithm is discussed.Finally,get the experiment data of the manipulator,then construct a least square model to estimate the parameters of joint incremental model.Combining the camera sampling rate to determine the discrete model of visual servo,analyze the model performance of the delay caused by different sampling rates of the visual sensor.The result of stable gain threshold is given,the control algorithm is analyzed and experimented,and the conclusion is drawn based on the general description of the servo result.