Hybrid Bio-Signal Control for Multi-DOF Prostheses
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing powered multi-functional limb prostheses struggle to achieve natural and precise movements across multiple degrees of freedom, with classification methods limiting simultaneous control and regression methods being restricted to a maximum of four functions corresponding to two degrees of freedom.
Innovation Solution
A multi-functional limb movement auxiliary device incorporating a base member, socket member, actuators, bio-signal sensing unit, and control unit that employs a combination of novelty detection and sequential/simultaneous estimators to enable articulated movements in at least two independent degrees of freedom, allowing for natural and precise control by comparing bio-signals with pre-determined thresholds and activating actuators accordingly.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If classification methods are used to control powered multi-functional limb prostheses, then articulation of many degrees of freedom is enabled, but arbitrary combinations of simultaneous activations across multiple DOFs are not allowed, limiting natural appearance of prosthetic movement
Solution Approach 1:
The patent applies dynamics by making the control method adaptable and changeable based on real-time conditions. The system dynamically switches between classification mode (for precise single-DOF control) and regression mode (for simultaneous multi-DOF control) depending on the complexity of the intended movement, allowing the prosthesis to naturally adapt to different movement scenarios and achieve both precision and natural appearance
Solution Approach 2:
The patent changes the control parameter regime by using novelty detection to identify whether the current bio-signal pattern represents a novel complex movement or a known simple movement. Based on this parameter change detection, the system switches between different control algorithms (classification vs. regression), enabling flexible adaptation between precise control and natural simultaneous multi-DOF movement
2Ease of operation
If regression methods are used to facilitate natural appearance with simultaneous and proportional control, then simultaneous movement estimation is enabled, but controlling capability is limited to a maximum of four functions corresponding to two DOFs
Solution Approach 1:
The patent segments the control space into two distinct regimes: simple single-DOF movements handled by classification methods and complex simultaneous multi-DOF movements handled by regression methods. This segmentation allows each method to operate within its optimal capability range, with the novelty detection system determining which regime applies at any given moment, thereby overcoming the DOF limitation of pure regression approaches
Solution Approach 2:
The patent creates a universal control system that can handle both simple and complex movements across any number of degrees of freedom. By integrating both classification and regression capabilities within a single hybrid framework, the system achieves multi-functionality that adapts to different movement complexities, eliminating the two-DOF restriction while maintaining natural appearance
3Ease of operation
If a hybrid control system combining novelty detection and sequential/simultaneous estimators is implemented, then natural and precise control in multiple DOFs is achieved, but device complexity increases
Solution Approach 1:
The patent applies preliminary action by pre-training the novelty detection system during a calibration phase with the user's specific movement patterns. This preliminary training enables the system to quickly and accurately identify whether current bio-signals represent novel complex movements or known simple movements, allowing rapid switching between control modes without adding real-time computational complexity to the operational phase
Data Source
AI summary
A method of operating a multi-functional limb movement auxiliary device comprising a plurality of actuators configured to move, upon activation, limb movement auxiliary device members in at least two independent degrees of freedom, a bio-signal sensing unit that is configured to acquire bio-signals indicative of motor activity, a control unit configured to receive and evaluate the acquired bio-signals, the method comprising steps ofacquiring bio-signals from the user of the limb movement auxiliary device,applying a novelty detection method (ND) for comparing digitally converted, acquired bio-signals with calibration bio-signals, and, based on a result of the comparison, for assigning a similarity measure to the acquired, digitally converted bio-signals,applying at least one sequential estimator (SEQ-E) to the acquired, digitally converted bio-signals, if the assigned similarity measure is equal to or larger than a pre-determined threshold value,applying at least one simultaneous estimator (SIM-E) to the acquired, digitally converted bio-signals, if the assigned similarity measure is less than the pre-determined threshold value, andactivating at least one actuator out of the plurality of actuators, based on a result of applying one of the estimators (SIM-E, SEQ-E) to the acquired, digitally converted bio-signals; anda multi-functional limb movement auxiliary device including a control unit configured for carrying out such method.