Gait Phase Recognition for Exoskeleton Control Using Adaptive Neuro Fuzzy Inference System

Abstract

Pressure measurement at the foot-ground contact zone provides necessary information for detecting the phases of the human gait. The human gait is a complicated process and, therefore, it is not possible to distinguish its stages simply by comparing the foot-ground surface pressure to threshold values. In this paper we propose a new method for human gait phases recognition using an adaptive network-based fuzzy inference system which processes data measured by pressure sensors located underneath footbed inserts. This method is intended to bring some additional information to the lower limb exoskeleton control system as well as for gait diagnostic purposes. The device uses only three pressure sensors for each foot. In order to verify the results of the study, measurements with sensors placed under the footbed inserts has been performed and the pressure at the foot-ground contact points during the walk has been investigated. Twelve models of gait phase classifiers were created. The performance of each classifier has been tested for a previously unknown pressure pattern that corresponds to the natural gait. Based on the quality of gait matching the test run, the best model of the classifier was selected.

Past, Present and Future of Assistive Robotic Lower Limb Exoskeletons

“…The devices which are existing now in the market are designed with various control systems and advanced algorithms with in the system. But there is a greater scope as well for improving the control algorithm with present Artificial intelligence [36].…”

Research on Upper-Body Exoskeletons for Performance Augmentation of Production Workers

“…Even though most studies on exoskeletons focus on very specific features, such as gait phase recognition, this research was performed in different fields [4]. Generally, an exoskeleton is a versatile device that can be used in various fields.…”