EDM Electrode Control via Iterative Learning

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Conventional electric discharge machining (EDM) process control methods result in bumpy movement of the controlled axis, significant electrode wear, and prolonged stabilization times after process pauses, due to their reliance on instantaneous feedback rather than historical data, leading to inefficiencies in material removal rate and surface quality.

Innovation Solution

Implementing an iterative learning control (ILC) method that uses historical data from previous machining cycles to adjust the tool electrode's movement, incorporating both deviation values from previous cycles and instantaneous process parameters to optimize the working gap distance and actuation parameters, thereby smoothing the axis movement and improving process stability.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Reliability

If conventional feedback control comparing current and desired working gap distance is used, then the process can be safely controlled, but the movement of the controlled axis becomes bumpy and stabilization time after process pauses increases

Engineering Contradiction:
Improveprocess control safetyVSAvoidmaterial removal rate
Core Design Contradiction:
ReliabilityVSProductivity

Solution Approach 1:

The patent applies preliminary action by using historical tracking error data from previous machining cycles to pre-adjust the tool electrode position before actual machining occurs. The iterative learning control algorithm processes past cycle errors and generates corrected command values in advance, allowing the system to anticipate and compensate for position deviations rather than reacting to them after they occur. This eliminates the need for conservative safe-distance approaches and reduces stabilization time after flushing motions.

Inventive Principle:
Principle #10Preliminary action

2Reliability

If conventional feedback control is used, then the process can be controlled, but valuable information from tracking error of each repetition is lost

Engineering Contradiction:
Improveprocess controlVSAvoidtracking error information
Core Design Contradiction:
ReliabilityVSLoss of information

Solution Approach 1:

The patent implements feedback by systematically collecting and processing tracking error data from each machining cycle. The iterative learning control algorithm stores the difference between desired and actual working gap positions from previous cycles and uses this feedback information to generate corrected command values for subsequent cycles. This creates a continuous learning loop where each cycle's errors inform the next cycle's performance, transforming lost information into valuable process improvement data.

Inventive Principle:
Principle #23Feedback

3Reliability

If frequent flushing motions are performed, then the working gap is cleared, but the time to get steady process condition after each pause increases

Engineering Contradiction:
Improveworking gap cleanlinessVSAvoidstabilization time
Core Design Contradiction:
ReliabilityVSLoss of time

Solution Approach 1:

The patent applies preliminary action by using historical tracking error data from previous machining cycles to pre-adjust the tool electrode position before actual machining occurs. The iterative learning control algorithm processes past cycle errors and generates corrected command values in advance, allowing the system to anticipate and compensate for position deviations rather than reacting to them after they occur. This eliminates the need for conservative safe-distance approaches and reduces stabilization time after flushing motions.

Inventive Principle:
Principle #10Preliminary action

4Reliability

If conservative approach distance is used after flushing motion, then electrode damage is avoided, but productivity decreases

Engineering Contradiction:
Improveelectrode protectionVSAvoidmaterial removal rate
Core Design Contradiction:
ReliabilityVSProductivity

Solution Approach 1:

The patent implements feedback by systematically collecting and processing tracking error data from each machining cycle. The iterative learning control algorithm stores the difference between desired and actual working gap positions from previous cycles and uses this feedback information to generate corrected command values for subsequent cycles. This creates a continuous learning loop where each cycle's errors inform the next cycle's performance, transforming lost information into valuable process improvement data.

Inventive Principle:
Principle #23Feedback

Data Source

PatentEP2179811B1Method and apparatus for controlling an electric discharge machining process
Publication Date: 2018.10.03 AGIE CHARMILLES SA
  • EP2179811B1 patent drawingFigure 1~2
  • EP2179811B1 patent drawingFigure 3~4b
  • EP2179811B1 patent drawingFigure 5

AI summary

Method for controlling an electric discharge machining process, wherein a tool electrode is moved relatively to a workpiece with a working gap distance, wherein the process comprises a current and at least one previous erosion cycle, the current and the previous erosion cycle each being divided into predetermined time intervals each comprising at least one discharge pulse, wherein similar working gap conditions are present within a time interval of the previous erosion cycle and of the current erosion cycle, and wherein subsequent erosion cycles are separated by a process pause cycle, the method comprising the steps of: measuring a value of a significant process parameter within a time interval of the previous erosion cycle, the significant process parameter being indicative of the working gap distance; determining a deviation value based on the measured value and a desired value of the significant process parameter of the time interval of the previous erosion cycle; and in the current erosion cycle, controlling the relative movement of the tool electrode in the erosion direction within the time interval of the current erosion cycle based on the deviation value determined for the time interval of the previous erosion cycle and at least one instantaneous process parameter being indicative of the instantaneous process conditions.