Development of a double inverted pendulum model for human gait analysis

• Main Author: Ibrahim, Zameera
• Format: Thesis
• Language: English
• Published: 2017
• Subjects: QH301 Biology
• Online Access: http://eprints.utm.my/id/eprint/87192/1/ZameeraIbrahimMSBME2018.pdf; http://eprints.utm.my/id/eprint/87192/; http://dms.library.utm.my:8080/vital/access/manager/Repository/vital:132594

Gait analysis is an essential tool in clinical rehabilitation. It facilitates the diagnosis, treatment, monitoring and implementation of methodologies that mitigate the effect of some pathologies associated with the movement. This analysis uses a systematic quantification, follow up and interpretation of the temporal sequence of movement that characterize human locomotion. It also requires a collection of kinematic and kinetic data that describes the displacements, angles and forces on the lower limbs and its joints during a gait cycle. However experimental procedures are difficult to carry out, time consuming and experienced people are required. The aim of this project is formulation and implementation of a double inverted pendulum (DIP) model to desrible the dynamics of human gait. The model consists the physiological extension of a head-arms-trunk (HAT) and leg segment. LaGrangeEular mechanics used for mathematical modelling. Simulations were performed,using simulink and MATLAB sortware. The hip angle, angle. ankle angle and ground reaction forces were compared against experimental data. It was found that experimental hip angle matched against each other from 20% to 75% of gait cycle while the experimental ankle angle deviated considerably from model ankle angle from 50% gait cycle onwards. In addition, GRF calculations showed that both model and experiment horizontal GRF reached the same maximum value of 15% BW while the minimum values deviated considerably. On the other hand, vertical GRF of experiment and model followed a similar pattern although significant deviation were observed at mid cycle. As a result, the computed Root Mean Square Errow (RMSE) for the modelled and experimental results were considerably higher which could be attributable to uncertainties in the model input parameters.