Genetic Optimization of a Manipulator: Comparison between Straight, Rounded, and Curved Mechanism Links

There are several ubiquitous kinematic structures that used in industrial robots, with the most prominent being a six-axis angular structure. However, researchers experimenting task-based mechanism synthesis could provide higher efficiency custom optimized manipulators. Many studies have focused on finding efficient optimization algorithm for robot These manipulators, however, usually from simple modular joints and links, without exploring more elaborate modules. Here, we show link modules defined by small numbers of parameters better performance than complicated ones. We compare four different manipulator types, namely basic predefined links fixed dimensions, straight can be lengths, rounded curvature Hermite spline. Manipulators then built these using genetic three tasks. The results demonstrate manipulators not only converge faster, which is expected given fewer parameters, but also lower cost values.