Wearable Robot Torque Estimation via Disturbance Observer
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing wearable robots face challenges in accurately determining the intended torque of a wearer without using expensive force-torque sensors, which are heavy and increase the weight and cost of the mechanism, and require real-time friction compensation for proper movement.
Innovation Solution
A method and system using a disturbance observer and an extended-Kalman filter to calculate the intended torque of a wearer by measuring the angle or angular velocity of a motor and a link connected to a joint, removing the need for a separate force-torque sensor, and compensating for friction without additional hardware.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If a force-torque sensor is attached to measure torque applied by a user, then measurement precision is improved, but device complexity and weight increase
Solution Approach 1:
The patent replaces the mechanical force-torque sensor with a computational approach using a disturbance observer. The observer calculates intended torque by processing signals from existing motor sensors and link sensors, substituting a complex mechanical measurement system with an information-processing system that uses mathematical models and signal filtering.
Solution Approach 2:
The disturbance observer acts as an intermediary between the existing sensors and the torque calculation. It processes signals from motor sensors and link sensors through mathematical operations (including Kalman filtering and disturbance estimation) to derive the intended torque, avoiding the need for direct mechanical torque measurement.
2Measurement precision
If a force-torque sensor is used to calculate wearer-intended torque, then measurement precision is improved, but weight increases
Solution Approach 1:
The patent eliminates the need for a physical force-torque sensor by replacing it with a computational disturbance observer that processes signals from existing sensors. This substitution removes the heavy sensor hardware while maintaining the ability to calculate intended torque through mathematical processing of motor and link signals.
3Stability of the object's composition
If friction compensation is implemented in real-time, then control stability is improved, but device complexity increases
Solution Approach 1:
The disturbance observer performs preliminary estimation of friction torque and other disturbances by processing sensor signals before they affect the motor control. The Kalman filter and disturbance estimation algorithms continuously predict and compensate for friction effects, preparing compensation values in advance rather than reacting after instability occurs.
Solution Approach 2:
The system implements feedback by continuously monitoring motor and link signals, comparing them against the disturbance observer's predictions, and adjusting the friction compensation in real-time. The disturbance estimation loop provides continuous feedback that refines the compensation accuracy, stabilizing control despite the added complexity of the feedback mechanism.
Data Source
AI summary
A method for extracting intended torque for a wearable robot includes a motor torque calculating step, a link rotation calculating step and an intended torque calculating step. In the motor torque calculating step, motor torque is calculated from the angular velocity of rotation of the motor. In the link rotation calculating step, the angular velocity of rotation of the link is calculated. In the intended torque calculating step, the motor torque and the angular velocity of rotation of the link are substituted into a disturbance observer, and an estimated value of the intended torque applied by a wearer is calculated.