Positioning Reliability Analysis Of Industrial Robots Through Probability And Evidence Theories

As one of the important equipment of intelligent manufacturing equipment,industrial robot has many functions,high speed and high precision.It has been widely used in various fields,typical applications include: assembly,welding,spraying and handling.In order to ensure product quality,improve production efficiency and ensure production safety,the positioning accuracy of industrial robots has become a common concern of scholars and engineers.At present,the repetitive positioning accuracy of industrial robots is high,while the absolute positioning accuracy is generally low.The factors that affect the absolute positioning accuracy include the errors in the production process,the errors in the assembly process,the wear of key components and the influence of the operating environment,etc.These factors are often uncertain.In order to study the influence of uncertainty in industrial robot system on its absolute positioning accuracy and improve the positioning accuracy reliability of industrial robot,this paper focuses on the coexistence of aleatory uncertainty and epistemic uncertainty in industrial robot system.The main work is as follows.(1)In view of the coexistence of aleatory uncertainty and epistemic uncertainty in industrial robot systems,a aleatory-epistemic hybrid positioning accuracy reliability analysis method for the industrial robots is proposed.Firstly,probability theory and evidence theory are used to describe the aleatory uncertain parameters and epistemic uncertain parameters respectively in the industrial robot system.Secondly,random variables and evidence variables are introduced into the kinematic modeling process of the robot to construct the performance functions of positioning accuracy,and aleatory-epistemic hybrid positioning accuracy reliability analysis model of industrial robot is established.Finally,based on the mixed probability and evidence reliability analysis framework,the probability and interval nesting algorithm is used to calculate the interval value of the positioning accuracy reliable probability of the industrial robot.And the effectiveness of the proposed method is verified by the example of an industrial robot.(2)For the existence of both random variables and evidence variables in input variables of the performance function,an improved probability-evidence hybrid reliability analysis method is proposed,which can improve the calculation efficiency of reliability under the premise of ensuring the calculation accuracy.Firstly,the evidence variables in the input variables are transformed into random variables by uniformity approach.Secondly,the Kriging model of performance function is constructed by combining the Latin hypercube sampling method and Kriging theory.Then,the Kriging model is substituted into the probability-evidence hybrid reliability analysis framework,and the probability interval of reliability is calculated.Finally,two numerical examples and an industrial robot example are given to verify the effectiveness of the proposed method.(3)Based on LabVIEW and MATLAB softwares,the reliability analysis platform for industrial robot positioning accuracy is developed.This platform introduces the reliability analysis algorithm for industrial robots positioning accuracy proposed in this paper,and has the interface of secondary development,which can easily import other analysis algorithms.The industrial robot positioning accuracy reliability analysis platform consists of four modules: user login and management module,reliability analysis module,data acquisition and storage module,and data playback module.