Neuro-fuzzy control of electroactive polymer actuators

Abstract

Electroactive polymer actuators, especially those based on polypyrrole (PPy), possess unique characteristics such as an ability to operate at the macro or micro scale, large force-to-weight ratio, biocompatibility, low cost and operation in aqueous and non-aqueous environments. Therefore, they are very suitable for the establishment of bio-mimetic devices, single-cell manipulators, robotics, prosthetics, and numerous biomedical applications. In this paper, we report on the neuro-fuzzy control of these actuators, which are typified by the tri-layer polypyrrole actuators considered in this paper, in order to improve their positioning accuracy and speed of response. We experimentally evaluated two model-free intelligent control strategies, which are fuzzy logic PD+I control and neuro-fuzzy Adaptive Neural Fuzzy Inference System (ANFIS) control. The performance of these intelligent controllers is compared to that of a conventional Proportional Integral Derivative (PID) controller. The experimental results demonstrated that they significantly outperformed the conventional PID controller with an improvement in rise time of at least 18 times and in settling time of at least 2 times. To the best of authors' knowledge, this is the first study to design and evaluate fuzzy logic PD+I and neuro-fuzzy ANFIS PD+I intelligent control methodologies for an important class of electroactive polymer actuators.