Dynamic Notch Filter Tuning for Multi-Axis Servo Oscillations

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Mechanical oscillations in multi-axis control systems, such as robots, due to backlash, compliant couplings, and flexing of mechanical linkages, lead to positioning errors and reduced throughput.

Innovation Solution

Implement a dynamic notch filter in each motor drive of the multi-axis control system, updating at a faster rate than the command update rate, to attenuate frequencies causing oscillations, using feedback signals to adjust the filter settings based on the system's pose and loading.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Productivity

If acceleration and deceleration rates are increased to improve throughput, then productivity increases, but mechanical oscillations increase causing positioning errors

Engineering Contradiction:
ImprovethroughputVSAvoidpositioning accuracy
Core Design Contradiction:
ProductivityVSManufacturing precision

Solution Approach 1:

The notch filter is made dynamic by continuously updating its center frequency based on real-time feedback from position sensors and knowledge of current robot pose and loading conditions. This allows the filter to adapt to changing operating conditions, maintaining effectiveness across different acceleration rates and robot configurations without sacrificing positioning accuracy.

Inventive Principle:
Principle #15Dynamics

Solution Approach 2:

The system changes the parameters of the notch filter (specifically the center frequency) based on operating conditions such as robot pose, loading, and desired acceleration rates. By dynamically adjusting these parameters, the system can maintain high throughput while compensating for oscillations that occur at different acceleration levels.

Inventive Principle:
Principle #35Parameter changes

2Manufacturing precision

If settling time is increased to reduce positioning errors from oscillations, then positioning accuracy improves, but productivity decreases

Engineering Contradiction:
Improvepositioning accuracyVSAvoidthroughput
Core Design Contradiction:
Manufacturing precisionVSProductivity

Solution Approach 1:

The system uses feedback from position sensors and knowledge of robot state to continuously monitor oscillations and dynamically adjust the notch filter parameters. This closed-loop approach allows the system to maintain positioning accuracy while minimizing settling time, as the filter adapts in real-time to dampen oscillations without requiring extended settling periods.

Inventive Principle:
Principle #23Feedback

Solution Approach 2:

By making the notch filter dynamic and adaptive rather than static, the system can respond quickly to oscillations as they occur, reducing the time needed to settle at target positions while maintaining accuracy. The filter evolves with the system state, providing continuous correction without requiring prolonged settling times.

Inventive Principle:
Principle #15Dynamics

3Manufacturing precision

If a static notch filter is used to reduce oscillations, then positioning accuracy improves, but the system cannot adapt to changing operating conditions

Engineering Contradiction:
Improvepositioning accuracyVSAvoidadaptability to operating conditions
Core Design Contradiction:
Manufacturing precisionVSAdaptability or versatility

Solution Approach 1:

The notch filter transitions from a static to a dynamic system that continuously updates its parameters based on real-time feedback. The center frequency and other filter characteristics are adjusted according to current robot pose, loading conditions, and operational state, enabling the system to maintain positioning accuracy across varying operating conditions rather than being optimized for a single fixed state.

Inventive Principle:
Principle #15Dynamics

Solution Approach 2:

The system dynamically changes the parameters of the notch filter based on operating conditions. By monitoring robot state and adjusting filter parameters accordingly, the system maintains adaptability to different payloads, positions, and operational modes while preserving positioning accuracy, something a static filter cannot achieve.

Inventive Principle:
Principle #35Parameter changes

Data Source

PatentUS12372951B2Dynamic command notch filter
Publication Date: 2025.07.29 ROCKWELL AUTOMATION TECH INC
  • US12372951B2 patent drawing
  • US12372951B2 patent drawing
  • US12372951B2 patent drawing

AI summary

A system and method for reducing mechanical oscillations in a multi-axis control system provides a first command for a dynamic notch filter at a first update rate to multiple motor drives. Each motor drive is operatively connected to a motor for an axis in the multi-axis control system. Each motor drive receives a second command for desired operation of the motor at a second update rate. Operation of the dynamic notch filter in each motor drive is changed as a function of the first command at the first update rate, and each motor drive generates a desired output voltage for desired operation of the motor at a third update rate. The third update rate is faster than the second update rate, the second command is passed through the dynamic notch filter to generate a filtered command, and the desired output voltage is generated as a function of the filtered command.