Design and Implementation of Human-Like Ambidextrous Robotic Arms for Cooperative Tasks

Abstract

In this paper an anthropomorphic robotic complex which consists of a robotic body with two robotic arms is designed for the purpose of achieving cooperative tasks. This research platform also aims to accelerate the creation and testing of control algorithms for humanlike anthropomorphic robots. An adaptive optimization algorithm based on differential evolution approach is proposed to solve the inverse kinematic problem. This algorithm facilitates finding a desirable solution easily and with less calculation cost compared to analytical and numerical methods, the thing that allows achieving realistic and efficient motion. This algorithm maintains balance between exploring new options and using current knowledge. Newton-Euler recursive approach is used for determining the robots dynamic properties. Then the required actuators are chosen, and stress analysis is conducted to evaluate the structural integrity of the robot and its capacity to withstand substantial loads. Adjustable backstepping fuzzy control method is used that is more robust to uncertainties, nonlinearities, and external disturbances and offers the robot the ability to perform movement in space, cooperate, and complete tasks in a coordinated manner. Ellipsoid intersection localizing method is used for collision avoidance. The selected methods excel at predicting the joints ’ movements required to attain desired poses with great accuracy and avoid collision. In a future paper, a study of the hands of this robot will be introduced. This work has the potential to be applied in tasks that are dangerous for humans, or in human-like tasks such as maintenance and manufacturing.