Evaluation of a mesh-based contact model for optimal control problems using automatic differentiation
In recent years, there has been a growing research interest in the field of musculoskeletal gait, with a focus on enhancing the walking ability of older adults and individuals with disabilities due to accidents or illnesses. This research endeavours to comprehend the physical behaviour of muscles, l...
| Autor: | |
|---|---|
| Tipo de documento: | dissertação |
| Data de publicação: | 2023 |
| País: | España |
| Recursos: | Universitat Politècnica de Catalunya (UPC) |
| Repositório: | UPCommons. Portal del coneixement obert de la UPC |
| Idioma: | inglês |
| OAI Identifier: | oai:upcommons.upc.edu:2117/386215 |
| Acesso em linha: | https://hdl.handle.net/2117/386215 |
| Access Level: | Acceso aberto |
| Palavra-chave: | Knee Biomechanics Genolls Biomècanica Àrees temàtiques de la UPC::Enginyeria biomèdica |
| Resumo: | In recent years, there has been a growing research interest in the field of musculoskeletal gait, with a focus on enhancing the walking ability of older adults and individuals with disabilities due to accidents or illnesses. This research endeavours to comprehend the physical behaviour of muscles, ligaments, and joints that contribute to this movement. The objective of this thesis is to examine and compare the various contact model approaches used to analyse and simulate the contact forces and moments arising on the contact of human joints, particularly in the knee and knee prosthesis. Additionally, the study aims to integrate tangential forces into the original model which only considers normal forces in pressure contact models, and to evaluate and analyse the resulting differences in a tracking problem. To accomplish the objectives of this study, an automatic differentiation tools were employed calculate derivatives to solve an optimal control problem (OCP) using the CasADi library in a Matlab program. Therefore, the model needed to be continuously differentiable. The contact models developed within the group were tested, the results show that the original model had moderate accuracy in predicting lateral knee contact force (��� 2 value of 0.46 and an RMSE of 259.7 N) and better accuracy in predicting medial knee contact force (��� 2 value of 0.62 and an RMSE of 132.6 N). The inclusion of tangential forces in the pressure contact model led to mixed results, such as improved performance in hip flexion, but a decrease in accuracy for predicting both lateral and medial knee contact forces. The study indicates the need for further improvement in mesh-based contact models for knee joint simulation, especially in the inclusion of tangential forces. |
|---|