Construction method of robot terminal performance prediction model based on ga-rbf network

Abstract

The invention discloses a construction method of a robot terminal performance prediction model based on a GA-RBF network, builds a shaft joint acquisition hardware platform for robot terminal data collection, uses the EtherCAT bus and a laser tracker as auxiliary tools for terminal testing, and trains GA-RBF respectively The input and output data acquisition method of the RBF network; through the bus to collect the position, speed and torque feedback of each joint in real time, the data is used as the input of the GA-RBF network, and the terminal data collected by the laser tracking coordinate measurement system is used as the GA-RBF network. The output is to train a robot terminal performance prediction model based on the GA-RBF network. The present invention greatly improves the acquisition accuracy of the axis joint servo pulse, and has greatly improved the application of the follow-up RBF network in the prediction of the end data and the calculation of the end parameter accuracy by the axis joint data DH model, and the research of high-precision data is even better close to the actual meaning.

Description

technical field [0001] The invention relates to the practical engineering application of an intelligent algorithm, in particular, the application of an intelligent algorithm based on a genetic algorithm optimization RBF neural network (GA-RBF) in the performance test of a robot terminal. Background technique [0002] For today's fierce competition in the robot industry (here we mainly refer to the six-degree-of-freedom robot arm), most robot manufacturers urgently need a complete robot testing solution in order to improve their core competitiveness. Robot-related tests have also emerged as the times require, and the error accuracy at the end of the robot can best reflect the overall performance of the robot, so the end error accuracy has always been a hot issue in robot testing. There are many factors that cause the end error of the robot: including errors caused by structural parameters, errors caused by motion variables; there are also unavoidable random errors such as ine...

AI Technical Summary

Problems solved by technology

For example, most robot companies in China stay at a relatively basic level in the testing of robots, and still generally use external high-precision ranging instruments, such as laser trackers to measure and analyze the absolute positioning accuracy and repeat positioning accuracy of robots. Systematic robot test specifications, and no more systematic test procedures

At the same time, it is impossible to test the accuracy of each robot using a laser tracker in actual work.

Images