Sensorless Motion Assist Device Torque Control
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current motion assist devices for elderly individuals, such as power assist suits, face challenges including cumbersome design, instability due to myoelectric sensor issues, and inability to provide versatile support for various body motions without causing discomfort or stress.
Innovation Solution
A motion assist device with a link structure, actuator, joint value measuring unit, target torque determining unit, torque measuring unit, and response control unit that determines and applies target torques to reduce joint load, cancel accelerations, and compensate for external torques, allowing for natural and sensitive support without myoelectric sensors.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If myoelectric sensors are used to detect user intentions, then the sensitivity and responsiveness of the motion assist device is improved, but the device becomes cumbersome and unreliable due to sensor attachment issues
Solution Approach 1:
The patent extracts and eliminates the myoelectric sensors from the system, replacing them with a sensorless control approach that uses only joint angle and torque measurements. This removes the attachment complexity and reliability issues while maintaining the ability to detect user intentions through passive dynamic interaction.
Solution Approach 2:
The patent replaces the electrical sensing system (myoelectric sensors) with a mechanical sensing approach that measures joint angles and torques. The user's intentions are detected through the mechanical interaction between the user's body and the assist device, eliminating the need for skin-contact electrical sensors.
2Ease of operation
If viscous resistance in gear portions is reduced to improve joint motion sensitivity, then the responsiveness to user intentions is improved, but the mechanical strength and durability of the joint units may be compromised
Solution Approach 1:
The patent implements feedback control that actively compensates for viscous resistance and other frictional effects in the gear portions. By measuring the actual joint torque and comparing it with the expected torque based on user intentions, the system can dynamically adjust the actuator output to counteract the viscous resistance, maintaining sensitivity without requiring mechanically perfect low-friction components.
3Use of energy by moving object
If a predetermined torque pattern is applied to assist walking, then the energy efficiency is improved, but the adaptability to various walking patterns and user intentions is reduced
Solution Approach 1:
The patent transitions from a static, predetermined torque pattern to a dynamic, real-time torque generation system. The assist torque is continuously adjusted based on the measured joint angle, joint velocity, and estimated user intentions, allowing the system to adapt to various walking patterns, speeds, and terrains while maintaining energy efficiency through passive dynamic assistance.
Solution Approach 2:
The system enables the user to self-regulate the assist torque by naturally interacting with the device's passive dynamic characteristics. The user's own movements and forces serve as the control input, eliminating the need for complex external control signals or pre-programmed patterns, and allowing automatic adaptation to different walking styles.
4Weight of moving object
If the link structure is designed to be lightweight for ease of wearing, then the comfort is improved, but the structural strength and support capability may be reduced
Solution Approach 1:
The patent employs counterbalancing mechanisms within the link structure where counterweights are strategically positioned to offset the gravitational force on the user's limbs. This allows the structure to be lightweight while still providing adequate support, as the counterweights create balancing moments that reduce the load on the structural components.
Data Source
AI summary
Intentions of a user are read from movements of the joints of the user without the use of a myoelectric sensor, and a force to support motions of the user is generated.Even if there is friction that is difficult to model in a gear portion of a joint actuator, the friction is compensated for, and joint units are controlled to follow an idealized mathematical model. In this manner, the force to support motions is generated, without giving any uncomfortable feeling to the joints of the user wearing the device. When motions of the user are supported, a mathematically-determined torque is applied to the joint units, so that a natural supporting force is constantly provided to the user in various circumstances.