Servo Control Output Visualization for Machine Learning Progress
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing output devices for servo control systems do not effectively allow users to monitor the progress or results of machine learning processes, particularly in servo control devices for machines and robots, as they fail to display learned parameters and evaluation function values, making it difficult for operators to grasp machine characteristics.
Innovation Solution
An output device that acquires and displays the relationship between learned parameters and evaluation function values, enabling operators to understand the machine learning progress and characteristics by showing the connection between these values and physical quantities, and allowing adjustments to be made to the servo control device parameters or search ranges.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If machine learning is performed on servo control device parameters, then the control performance is improved, but the operator cannot check the progress or result of machine learning
Solution Approach 1:
The patent implements a feedback mechanism by displaying the evaluation function value and learned parameters on a display unit. This allows operators to monitor the machine learning progress in real-time, receiving visual feedback about the optimization process while maintaining improved control performance through the learned parameters.
Solution Approach 2:
The patent introduces an intermediary display system that mediates between the machine learning process and the operator. The display unit acts as an intermediary, translating complex machine learning data (evaluation function values, parameters) into visual information that operators can understand and monitor, without interfering with the actual control performance improvement.
2Loss of information
If evaluation function value is displayed, then machine learning progress becomes visible, but operator has difficulty grasping machine characteristic from the evaluation function value
Solution Approach 1:
The patent introduces physical quantity information as an intermediary that bridges the gap between the abstract evaluation function value and the operator's understanding of machine characteristics. This intermediary layer translates complex optimization data into more intuitive physical terms that operators can readily comprehend.
Solution Approach 2:
The patent uses visual display methods to present information in an operator-friendly format. By displaying both evaluation function values and physical quantity information on the display unit, the system makes the data visually accessible and easier to interpret, though the specific color coding mechanism is not detailed in the patent text.
3Reliability
If parameters of constituent element are machine learned, then servo control performance is improved, but the parameter and evaluation function value are not displayed
Solution Approach 1:
The patent implements feedback by displaying both the learned parameters of constituent elements and their corresponding evaluation function values. This allows operators to see the direct relationship between parameter optimization and performance improvement, maintaining transparency while achieving enhanced servo control performance.
Solution Approach 2:
The display unit serves multiple functions: it displays evaluation function values, learned parameters, and physical quantity information simultaneously. This multi-functional display system addresses the information display need without adding separate dedicated devices, maintaining system efficiency while improving transparency.
Data Source
AI summary
An object is to make it possible to acquire a parameter or a first physical quantity that has been learned and an evaluation function value and thereby check the progress or the result of machine learning. An output device includes: an information acquisition unit which acquires, from a machine learning device that performs machine learning on a servo control device for controlling a servo motor driving the axis of a machine tool, a robot or an industrial machine, a parameter or a first physical quantity of a constituent element of the servo control device that is being machine learned or has been machine learned and an evaluation function value; and an output unit which outputs information indicating a relationship between the acquired parameter, the first physical quantity or a second physical quantity determined from the parameter and the evaluation function value.