| CPC G06N 20/00 (2019.01) [G06N 5/022 (2013.01)] | 7 Claims |
|
1. A data processing system for accurately predicting a motion trajectory of an industrial robot moving between a given pair of robotic locations in a robotic simulation system, the data processing system comprising:
a processor; and
an accessible memory connected to said processor, said processor being configured to:
a) receive training data of motion trajectories of the robot for a plurality of robotic location pairs, and increase an amount of training data sets in the training data, wherein the training data that was received includes an original robotic location pair having an original start location and an original target location, wherein the amount of training data sets in the training data is increased by adding at least one training data set that includes a start location and a target location, and wherein the start location is an intermediate location positioned between the original start location and the original target location, and the target location is the original target location from the original robotic location pair;
b) process the training data so as to obtain x tuples and y tuples for machine learning purposes; wherein the x tuples describe the robotic location pair and the y tuples describe one or more intermediate robotic locations with poses at specific time stamps during the motion of the robot between the location pair;
c) learn from the processed data a function mapping the x tuples into the y tuples so as to generate a motion prediction module for mimicking the kinematics motion behavior of the robot; and
d) for a given robotic location pair, predict the robotic motion between the given pair by getting the corresponding intermediate locations resulting from the motion prediction module.
|