Robot Learning Control Parameter Tuning for Vibration Damping

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

The existing learning control units for robots, generated before shipment, do not account for the specific operating conditions at the user's end, leading to suboptimal vibration reduction during production.

Innovation Solution

A controller with a learning control unit that adjusts parameters based on frequency response characteristics at the time of production, using a parameter storage unit to set optimal parameters for the learning control unit, enhancing vibration reduction effectiveness.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Reliability

If a learning control unit is generated before robot shipment based on frequency response characteristics in a standard load state, then the robot has initial vibration reduction capability, but the learning control unit cannot achieve optimum vibration reduction effect at the user's end under different operating conditions

Engineering Contradiction:
Improvevibration reduction effectVSAvoidadaptability to different operating conditions
Core Design Contradiction:
ReliabilityVSAdaptability or versatility

Solution Approach 1:

The learning control unit parameters are made dynamically adjustable rather than fixed. The system allows parameter updates at the user's end based on actual operating conditions, transforming the static pre-configured controller into a dynamic adaptive system that can optimize vibration reduction for different tools, postures, and production scenarios

Inventive Principle:
Principle #15Dynamics

Solution Approach 2:

The invention changes the parameters of the learning control unit based on frequency response characteristics measured under actual operating conditions. By adjusting parameters such as learning gain and filtering characteristics according to the specific robot configuration and production environment, the system achieves optimal vibration reduction performance for each unique operating scenario

Inventive Principle:
Principle #35Parameter changes

2Device complexity

If the learning control unit parameters are fixed before shipment, then the device complexity is reduced, but the manufacturing precision and path accuracy deteriorate under different production conditions

Engineering Contradiction:
Improvecontrol system complexityVSAvoidpath accuracy
Core Design Contradiction:
Device complexityVSManufacturing precision

Solution Approach 1:

The system performs preliminary configuration of the learning control unit with default parameters before shipment, enabling basic vibration reduction functionality out of the box. This preliminary setup reduces initial complexity while maintaining the capability for later parameter optimization to achieve high path accuracy under specific production conditions

Inventive Principle:
Principle #10Preliminary action

Solution Approach 2:

The system enables self-service parameter adjustment at the user's end through automated frequency response measurement and parameter optimization routines. This allows the control unit to adapt to specific operating conditions without requiring complex manual tuning or external assistance, balancing simplicity with precision

Inventive Principle:
Principle #25Self-service

Data Source

PatentUS20240033909A1Control device, robot control device, and control method
Publication Date: 2024.02.01 FANUC LTD
  • US20240033909A1 patent drawing
  • US20240033909A1 patent drawing
  • US20240033909A1 patent drawing

AI summary

The present invention improves a vibration-dampening effect produced through learning in a state during production on a user side. This control device, which prepares a correction amount for controlling the operations of a robot, comprises: a learning control unit that has a parameter used in a learning control for preparing the correction amount; a parameter storage unit that stores a parameter set prior to shipment; and a parameter adjustment unit that, during production by the robot, adjusts the parameter stored by the parameter storage unit and sets the adjusted parameter in the learning control unit. The parameter adjustment unit adjusts the parameter on the basis of, e.g., the multiplicative inverse of a frequency response characteristic of the robot. The parameter adjustment unit also adjusts the parameter according to, e.g., a genetic algorithm.