Servo Control Evaluation Function Output for Transparent Machine Learning
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Users cannot effectively suppress errors in servo control systems, as existing technologies lack clear visualization of evaluation functions and learning results, leading to inconsistent performance in machine learning applications for servo motors in machine tools and robots.
Innovation Solution
An output device that displays and outputs multiple evaluation functions and their corresponding machine learning results, allowing users to select and change evaluation functions based on intended learning outcomes, thereby improving error suppression and system performance.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Extent of automation
If machine learning is performed using reinforcement learning to adjust control parameters, then the control parameter adjustment can be automated, but the user cannot understand which learning effect is obtained when a certain evaluation function is used
Solution Approach 1:
The patent implements feedback by displaying the evaluation function values and learning results to the user. The control device outputs the evaluation function value calculated by the machine learning device and the learning result, allowing the user to see the relationship between the evaluation function and the learning effect. This feedback mechanism resolves the contradiction by maintaining automation while providing transparency through information feedback.
2Device complexity
If a single evaluation function is used for machine learning, then the learning process is simple, but different users have different requirements for error suppression and vibration suppression that cannot be simultaneously satisfied
Solution Approach 1:
The patent applies dynamics by allowing the evaluation function to be dynamically changed based on user requirements. The control device includes a storage unit that stores multiple types of evaluation functions, and the machine learning device can switch between different evaluation functions according to the user's needs. This enables the system to adapt to different user requirements (error suppression vs. vibration suppression) while maintaining a relatively simple learning process structure.
3Adaptability or versatility
If multiple evaluation functions are used to satisfy different user requirements, then user-specific performance can be optimized, but the system complexity increases
Solution Approach 1:
The patent implements universality by designing a control device that can handle multiple evaluation functions through a unified architecture. The storage unit stores multiple types of evaluation functions, and the machine learning device can select and use different evaluation functions based on user requirements. This multi-functional design allows the system to satisfy different user requirements (error suppression, vibration suppression, or both) without requiring separate systems for each case, thus managing complexity while maintaining versatility.
Data Source
AI summary
A plurality of evaluation functions and a machine learning result of each of the evaluation functions are output so that a relation between the evaluation function and the learning result can be ascertained. An output device includes: an output unit that outputs a plurality of evaluation functions used by a machine learning device that performs machine learning of parameters of components of a servo control device that controls a servo motor that drives an axis of a machine tool, a robot, or an industrial machine and a machine learning result of each of the evaluation functions; and an information acquisition unit that acquires the machine learning result from at least one of the servo control device and the machine learning device.