An Immune Evolutionary Algorithm Based Pose Estimation Method for Parallel Manipulator

Based on immune systems, a new immune evolutionary algorithm (LEA) is presented to develop a pose estimation method for a parallel manipulator in the paper. Four vertices of a parallelogram device on a parallel manipulator's end-effector are used as the object model. And the problem of pose identification is transformed to obtain the optimal depth estimations of the object model. In IEA, depth estimations of the object model are taken as an antigen. Then the optimal solutions are searched by clone selection and variation operator. In theory, this method enriches the pose estimation methods from four points correspondences. In addition, it provides guidance for practical applications of a parallel manipulator. Experiments results demonstrate that our algorithm works speedily and robustly.