| The hypothesis upon which our study rests is that a massive array acting in parallel can be regarded as a simplified model of biological muscle. In the study of biomechanics, it has been found that the control strategy used in the locomotion of a biological system is mainly via muscle fiber recruitment. In other words, each muscle fiber is postulated to operate in a binary fashion.; Bundling large numbers of active elements into a limited space is becoming feasible with recent developments in smart materials. The study presented in this dissertation is not directed at the development of actuator array elements but at the use of these active elements for robotic applications. We propose methods to characterize, control and coordinate systems actuated by large numbers of binary actuators. Nevertheless, our work may also influence the development of smart materials by providing attractive applications of such elements and identifying desirable characteristics which may be built into them. |
