Servo Drive Machine Learning for Analog Command Error Correction
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing drive apparatuses face errors in interpreting analog voltage commands due to input offset voltage and noise, leading to decreased controllability, responsiveness, and increased power consumption, with conventional solutions being inadequate in addressing these issues.
Innovation Solution
A drive apparatus equipped with a machine learning apparatus that learns to optimize commands by observing external commands, state variables, and determination data, using a multilayered structure to improve conversion accuracy and efficiency.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If conventional measures such as dead band or filter are provided to suppress error in received analog voltage command value, then error suppression is improved, but command value fluctuations below threshold and errors exceeding threshold cannot be accommodated
Solution Approach 1:
The patent implements feedback by having the drive apparatus send back information about the received command value to the control apparatus. This allows the control apparatus to detect errors in command interpretation and send correction commands, thereby maintaining both error suppression and adaptability to different error conditions.
Solution Approach 2:
The patent introduces an intermediary communication mechanism between the drive apparatus and control apparatus. The drive apparatus acts as an intermediary that not only receives commands but also provides feedback about command interpretation accuracy, enabling the control apparatus to adjust its command output accordingly.
2Reliability
If elaborate mechanism including addition of hardware and complex control by control apparatus is constructed to prevent servo motor malfunction, then reliability is improved, but device complexity increases
Solution Approach 1:
The drive apparatus performs self-service by autonomously detecting errors in received command values and generating correction commands without requiring complex external hardware or control mechanisms. This reduces device complexity while maintaining reliability through the drive apparatus's own error detection and correction capabilities.
Solution Approach 2:
The feedback mechanism allows the drive apparatus to communicate error information back to the control apparatus, enabling error prevention without adding elaborate hardware. The feedback loop provides a simple yet effective way to maintain reliability while minimizing complexity.
3Ease of operation
If analog voltage command is transmitted to control servo motor, then command transmission is achieved, but error occurs between intended command value and interpreted command value due to input offset voltage, noise, and temperature variation
Solution Approach 1:
The feedback mechanism enables the drive apparatus to report back the actual received command value to the control apparatus. This allows the control apparatus to detect discrepancies between intended and received values caused by analog transmission errors and send appropriate correction commands, thereby maintaining command value accuracy.
Solution Approach 2:
The patent replaces the purely analog voltage transmission system with a hybrid system that incorporates digital communication for feedback. This substitution allows for more accurate command transmission by combining analog actuation with digital error detection and correction.
Data Source
AI summary
A drive apparatus which controls a speed or a torque of a servo motor of a processing machine according to a command obtained by converting an external command input from outside includes a machine learning apparatus which learns appropriate conversion from the external command to the command. The machine learning apparatus includes: a state observing unit which observes the external command, the command, and a state of the processing machine or the servo motor as state variables that represent a present state of an environment; a determination data acquiring unit which acquires determination data indicating an evaluation result of processing by the processing machine; and a learning unit which learns the external command and the state of the processing machine or the servo motor, and the command, in association with each other using the state variables and the determination data.