Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

Design of Fuzzy Logic Parameter Tuners for Upper-Limb Assistive Robots

Christopher Coco, Jonathan Spanos, Hamid Osooli, Reza Azadeh

TL;DR

Manual tuning of high-level device gains in upper-limb assistive robotics is tedious and task-dependent. The study designs two Takagi-Sugeno-Kang fuzzy controllers that adjust the dominant and opposing gains in real time based on user intent and muscle activation, implemented on the MyoPro 2. The controllers rely on expert-defined rules and Gaussian membership functions within a TS-K framework, and are evaluated across three tasks showing reductions in user–device fighting measured by pressure sensors. The results suggest that task-adaptive gain tuning via fuzzy logic can improve user experience without requiring extensive training data, supporting more practical deployment of assistive exoskeletons.

Abstract

Assistive Exoskeleton Robots are helping restore functions to people suffering from underlying medical conditions. These robots require precise tuning of hyper-parameters to feel natural to the user. The device hyper-parameters often need to be re-tuned from task to task, which can be tedious and require expert knowledge. To address this issue, we develop a set of fuzzy logic controllers that can dynamically tune robot gain parameters to adapt its sensitivity to the user's intention determined from muscle activation. The designed fuzzy controllers benefit from a set of expert-defined rules and do not rely on extensive amounts of training data. We evaluate the designed controllers with three different tasks and compare our results against the manually tuned system. Our preliminary results show that our controllers reduce the amount of fighting between the device and the human, measured using a set of pressure sensors.

Design of Fuzzy Logic Parameter Tuners for Upper-Limb Assistive Robots

TL;DR

Manual tuning of high-level device gains in upper-limb assistive robotics is tedious and task-dependent. The study designs two Takagi-Sugeno-Kang fuzzy controllers that adjust the dominant and opposing gains in real time based on user intent and muscle activation, implemented on the MyoPro 2. The controllers rely on expert-defined rules and Gaussian membership functions within a TS-K framework, and are evaluated across three tasks showing reductions in user–device fighting measured by pressure sensors. The results suggest that task-adaptive gain tuning via fuzzy logic can improve user experience without requiring extensive training data, supporting more practical deployment of assistive exoskeletons.

Abstract

Assistive Exoskeleton Robots are helping restore functions to people suffering from underlying medical conditions. These robots require precise tuning of hyper-parameters to feel natural to the user. The device hyper-parameters often need to be re-tuned from task to task, which can be tedious and require expert knowledge. To address this issue, we develop a set of fuzzy logic controllers that can dynamically tune robot gain parameters to adapt its sensitivity to the user's intention determined from muscle activation. The designed fuzzy controllers benefit from a set of expert-defined rules and do not rely on extensive amounts of training data. We evaluate the designed controllers with three different tasks and compare our results against the manually tuned system. Our preliminary results show that our controllers reduce the amount of fighting between the device and the human, measured using a set of pressure sensors.
Paper Structure (9 sections, 3 equations, 5 figures, 4 tables)

This paper contains 9 sections, 3 equations, 5 figures, 4 tables.

Table of Contents

  1. Introduction
  2. Device Overview
  3. Fuzzy Controller Design
  4. Experiments
  5. Experimental Setup
  6. Task Design
  7. Data Collection & Pressure Sensors
  8. Results
  9. Conclusions & Future Work

Figures (5)

  • Figure 1: MyoPro 2 orthosis used in our research. (a) integrated sEMG sensors and (b) integrated pressure sensors for the arm joint.
  • Figure 2: Designed membership functions for the two controllers.
  • Figure 3: Control surfaces for the controllers: (left) $f_{E_d}$, (right) $f_{E_o}$.
  • Figure 4: Three tasks designed for evaluation. From left to right: horizontal motion, vertical motion, and pushing motion.
  • Figure 5: Recorded pressure sensors signals for the sensitive (top row), normal (middle row), and resistant scenarios (bottom row). The plots correspond to the vertical (left column), horizontal (middle column), and pushing tasks (right column). In each task, the joint angle trajectory is plotted. The vertical dashed lines indicate when one trial of each task was completed and the next one started.