Rule Based Intent Recognizer in Sit to Stand Movement Based on Knee Angle and Ground Reaction Force

Abstract

This paper presents an approach to detect phase transition from sit to stand (STS) movement in able bodied person based on knee angle and ground reaction force (GRF) data. Phase transition detection method developed in this paper is a heuristic rule based algorithm for the purpose of prosthetic knee device that have the limitation of number of parameters that can be obtained in the amputee. A single able-bodied subject experimental design is performed to obtained necessary data to construct the algorithm such as kinematics and kinetics data. Event detection is divided into three parts, namely start of STS event, seat-off event, and end of STS event, with its own rule that validated five-fold. From the ten performed trial, mean and standard deviation of STS time are 1.34 s and 0.29 s respectively and rule based algorithm detect the seat-off event within 10 ms early from the true seat-off event measured by the sit force plate.