Spindle Torque Control for Stable Tapping Deceleration
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional machine tool control devices face challenges in stabilizing motor ability during deceleration, leading to inconsistent tapping operations and extended cycle times, as the spindle torque command value in deceleration often exceeds target values.
Innovation Solution
A machine learning device that performs reinforcement learning to optimize the ratio of movement distances in acceleration and deceleration, using a reward-based system to update action value functions, thereby adjusting the torque command values to approximate the target torque command value in deceleration, ensuring stable tapping operations and reduced cycle times.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If the spindle torque command value in deceleration is increased to reduce deceleration period and extend constant speed period, then cycle time is reduced, but the spindle torque command value exceeds the target spindle torque command value in deceleration
Solution Approach 1:
The patent applies dynamics by making the torque command value adjustable and adaptive during deceleration. The system dynamically modifies the torque command value based on learned parameters specific to each machine and operation condition, transitioning from static conventional control to dynamic adaptive control. This allows the system to optimize cycle time while maintaining torque values within target ranges through real-time parameter adjustment.
Solution Approach 2:
The patent implements parameter changes by modifying the torque command value parameter during deceleration based on machine-specific and condition-specific characteristics. The machine learning component learns optimal parameter settings that reduce cycle time while keeping torque values within acceptable ranges. This involves changing the torque command value from fixed conventional settings to variable optimized settings tailored to each machine and operation condition.
2Productivity
If the spindle torque command value in deceleration is increased to reduce deceleration period, then cycle time is reduced, but tapping operations become inconsistent across different machines and operation conditions
Solution Approach 1:
The patent applies local quality by tailoring the torque command value parameters to specific machines and operation conditions rather than using universal fixed values. The machine learning component learns and stores optimal parameters for each machine and condition combination, allowing localized optimization. This ensures that each machine operates with parameters suited to its specific characteristics, achieving both reduced cycle time and consistent tapping operations across different machines.
Solution Approach 2:
The patent implements parameter changes by adapting the torque command value parameters based on machine-specific and condition-specific characteristics. The system learns optimal parameter settings for each machine and operation condition, enabling the same control strategy to achieve consistent results across different machines. This involves changing from fixed universal parameters to adaptive machine-specific parameters that maintain operation stability while improving productivity.
3Productivity
If the spindle torque command value in deceleration is increased to extend constant speed period, then cycle time is reduced, but the torque command value exceeds target values leading to control instability
Solution Approach 1:
The patent applies feedback by using machine learning to learn from operational data and adjust torque command values accordingly. The system continuously monitors operation outcomes and refines its parameter settings to maintain stability. This feedback mechanism allows the system to reduce cycle time through extended constant speed periods while preventing torque values from exceeding target ranges that would cause control instability.
Solution Approach 2:
The patent implements dynamics by making the torque command value adaptive rather than fixed. The system dynamically adjusts parameters based on learned machine characteristics and operation conditions, allowing flexible optimization of cycle time while maintaining control stability. This dynamic approach enables the system to extend constant speed periods for productivity improvement while automatically preventing torque values from exceeding stable operating ranges.
Data Source
AI summary
A machine learning device includes: a state information acquisition unit configured to cause the control device to execute a tapping program to acquire from the control device, state information including a torque command value with respect to the spindle motor, a drive state including deceleration, a ratio of a movement distance in acceleration and a movement distance in deceleration; an action information output unit configured to output action information including adjustment information of the ratio of the movement distance in acceleration and the movement distance in deceleration, to the control device; a reward output unit configured to output a reward value in reinforcement learning based on a torque command value in deceleration, and a target torque command value in deceleration; and a value function update unit configured to update an action value function based on the reward value, the state information, and the action information.