Optimization of a 2 DOF Micro Parallel Robot Using Genetic Algorithms

Over the last couple of decades parallel robots have been increasingly studied and developed from both a theoretical viewpoint as well as for practical applications (Merlet, 1995). Advances in computer technology and development of sophisticated control techniques have allowed for the more recent practical implementation of parallel manipulators. Some of the advantages offered by parallel manipulators, when properly designed, include an excellent load-to-weight ratio, high stiffness and positioning accuracy and good dynamic behavior (Merlet 1995, Stan 2003). The ever-increasing number of publications dedicated to parallel robots illustrated very well this trend. However, there are also some disadvantages associated with parallel manipulators, which have inhibited their application in some cases. Most serious of these is that the particular architecture of parallel manipulators leads to smaller manipulator workspaces than their serial counterparts. One of the main influential factors on the performance of the micro parallel robot is its structural configuration. The kinematic relations, statics, dynamics and structure stiffness are all dependent upon it. After its choice, the next step on the manipulator design should be to establish its dimensions. Usually this dimensioning task involves the choice of a set of parameters that define the mechanical structure of the parallel robots. The parameter values should be chosen in a way to optimize some performance criterion, dependent upon the foreseen application. Micro parallel robots can also be difficult to design (Stan, 2003), since the relationships between design parameters and the workspace, and behavior of the manipulator throughout the workspace, are not intuitive by any means. This is one of the reasons why Merlet (1995) argues that customization of parallel manipulators for each application is absolutely necessary in order to ensure that all performance requirements can be met by the manipulator. As a result, development of design methodologies for such manipulators is an important issue in order to ensure performance to their full potential. In particular, the development and refinement of numerical methods for workspace determination of various parallel manipulators is of utmost importance. There is a strong and complex link between the type of robot’s geometrical parameters and its performance. It’s very difficult to choose the geometrical parameters intuitively in such a way as to optimize the performance. Several papers have dealt with parallel robots to optimize performances (Stan, 2006). For example, various methods to determine workspace