Experimental characterization and validation of model for cantilever beam platform using SMA actuator

作     者：Santana Barroso, Antonio Julio de Moura Fernandes, Eisenhawer Da Rocha Neto, Jose Sergio 

作者机构：Ctr Univ Estacio Sergipe Aracaju Brazil Univ Wisconsin Madison UW Madison Madison WI USA Univ Fed Campina Grande UFCG Ctr Ciencias & Tecnol CCT Dept Engn Mecan DEM Campina Grande Paraiba Brazil Univ Fed Campina Grande Programa Posgrad Engn Mecan PPGEM Campina Grande Paraiba Brazil Univ Fed Campina Grande Programa Posgrad Engn Eletr PPgEE Campina Grande Paraiba Brazil Univ Fed Campina Grande Campina Grande Paraiba Brazil 

出 版 物：《IEEE LATIN AMERICA TRANSACTIONS》 (IEEE. Lat. Am. Trans.)

年 卷 期：2022年第20卷第3期

页      面：519-526页

核心收录：

主　　题：Computational modeling Deformable models Biological system modeling Actuators Metals Structural beams IEEE transactions Shape-memory alloys SMA actuator Simulation model Experimental characterization 

摘      要：Shape-memory alloys are being used and studied in different areas due to features such as feasibility, high deformation per volume ratio with shape recovery, soft and silente movements. Although these actuators present some advantages for use, its application in position control of structures has some drawbacks related to parameter characterization and, consequently, for system modeling. The lack of information for model the SMA actuator has been compensated in literature by using control algorithms such as Fuzzy controllers, RNA and Fuzzy-PID controller that requires hardware for implementation with more computational requirements. This work presents a model for a flexible cantilever beam actuated by shape-memory alloy (SMA) actuator. The SMA actuator is actived in order to control the position of the free tip of the cantilever beam. The proposed model in Simulink/Matlab is composed by mechanical and electrical sub-systems of the structure. The proposed model is validated by the comparison with the experimental results obtained from a laboratory test bench. The metric used to validate the model is Normalized Mean Squared Error (NMSE), reaching a value of 0.9981, and Normalized Root Mean Squared Error (NRMSE) of 0.9562, for a 90% duty cycle. As a result, the proposed model can be used for implementation of position closed-loop control of the structure based on conventional algorithms (PI or PID) suitable for simpler hardware presenting limited computational resources (microcontrollers).