Prediction of Ground Reaction Forces and Center of Pressure During Different Lifting Tasks Using Inverse Dynamics and Machine Learning Methods

Please enable javascript in your browser.

Daroudi, Sajjad | 2023

76 Viewed

Type of Document: M.Sc. Thesis
Language: Farsi
Document No: 56257 (08)
University: Sharif University of Technology
Department: Mechanical Engineering
Advisor(s): Arjmand, Navid
Abstract:
Ground reaction forces (GRFs) and their centers of pressure (CoPs) are essential inputs in musculoskeletal models of the spine that study load-lifting tasks. While GRFs and CoPs are usually measured in vivo via force-plates, they can be predicted based on motion equations, developed in AnyBody Modelling System (AMS) thus eliminating the need for laboratory equipment and in vivo measurements. This study aims to develop an easy-to-use yet accurate model that predict GRFs and CoPs during different Manual Material Handling (MMH) activities using AnyBody Modelling System and artificial neural networks. First, the accuracy of the AMS GRF prediction algorithm was evaluated for 8 normal- and 4 over-weight (BMI>25 kg/m2) individuals during 48 different lifting tasks. GRF and CoP values were predicted by AMS (v.7.4) driven by in vivo motion data. Subject-specific bottom-top OpenSim (v.3.2) musculoskeletal models were also driven using both the predicted and measured GRFs/CoPs to compare the predicted L5-S1 loads as AMS uses a top-down approach to predict spinal loads. Secondly, 1836 MMH activities for ten healthy males (20
Keywords:
Ground Reaction Force (GRF) ; Artificial Neural Network ; Manual Material Handeling ; Anybody Software ; Human Movment Analysis ; Inverse Dynamics ; Machine Learning

No TOC