GA-based control of binary-continuous joints of a multi-link micromanipulator using SMA actuators

The paper presents a control strategy for a hyper redundant micromanipulator that has the particularity to have either a) binary joints: only a small finite number of position are allowed within min and max bounds and b) continuous joints: the joint can move continuously within its min and max bounds. The problem optimization is double: a) solve inverse kinematics for this kind of system and b) maximize the number of binary joints. We show that the hybrid nature of the manipulator make these problems nontrivial and we develop a parallel evolution strategy (ES) and genetic algorithm (GA) in response to this dual problem.