Intelligent active force control of human hand tremor using smart actuator

Patients suffering from Parkinson’s disease (PD) experience tremor which may generate a functional disability impacting their daily life activities. In order to provide a non-invasive solution, an active tremor control technique is proposed to suppress a human hand tremor. In this work, a hybrid controller which is a combination of the classic Proportional-Integral (PI) control and Active Force Control (AFC) strategy was employed. A test-rig is utilized as a practical test and verification platform of the controller design. A linear voice coil actuator (LVCA) was utilized as the main active suppressive element to control the tremor of hand model in collocation with the sensor. In order to validate the AFC scheme in real-time application, an accelerometer was used to obtain the measured values of the parameter necessary for the feedback control action. Meanwhile, a laser displacement sensor was used to quantify the displacement signal while hand shaking. To optimize the controller parameters, three different optimization techniques, namely the genetic algorithm (GA), particle swarm optimization (PSO) and differential evolution (DE) techniques were incorporated into the hybrid PI+AFC controller to obtain a better performance in controlling tremor of the system. For the simulation study, two different models were introduced to represent the human hand in the form of a mathematical model with four degree-of-freedom (4 DOF) biodynamic response (BR) and a parametric model as the plant model. The main objective of this investigation is to optimize the PI and AFC parameters using three different types of intelligent optimization techniques. Then, the parameters that have been identified were tested through an experimental work to evaluate the performance of controller. The findings of the study demonstrate that the hybrid controller gives excellent performance in reducing the tremor error in comparison to the classic pure PI controller. Based on the fitness evaluation, the AFC-based scheme enhances the PI controller performance roughly around 10% for all optimization techniques. Besides that, an intelligent mechanism known as iterative learning control (ILC) was incorporated into the AFC loop (called as AFCAIL) to find the estimated mass parameter. In addition, a sensitivity analysis was presented to investigate the performance and robustness of the voice coil actuator with the proposed controller in real-time environment. The results prove that the AFCAIL controller gives an excellent performance in reducing the hand tremor error in comparison with the classic P, PI and hybrid PI+AFC controllers. These outcomes provide an important contribution towards achieving novel methods in suppressing hand tremor by means of intelligent control