Developments on Modeling for a Micro-Hand Based on Grid Search Optimized SVR Method

In recent years, soft actuators have been widely used in various fields, and most of them are flexible materials, among of which, the micro-hand is driven by pneumatic pressure. That is, when different amounts of pressure are applied to the rubber tube, the rubber tube will bend to different degrees, resulting in displacement in different directions. In previous studies, a microhand model was proposed, however, the error between the model and the real system was large. Therefore, we refer to the support vector regression method to improve the modelling accuracy based on the experimental data, which is good at dealing with small sample data. Moreover, the grid search method is applied to optimize the parameters arising from support vector regression. The simulation results show that the model constructed by using the modeling method of grid search optimized support vector regression is more accurate in estimating the input-output relationship of micro-hand compared with the previous model, Thus, the effectiveness of grid search optimized support vector regression modeling method is confirmed.