High Accuracy Target Tracking System Based on Muscle-skeleton Robotic Arm

Abstract

Human biological structures and neural mechanisms have provided many great inspirations for the design and improvement of human-like robots. McKibben muscle, as a new type of flexible actuator, can be used to manufacture humanoid robots. In this paper, a target tracking system based on muscle-skeleton robotic arm and MLP-driven pressure prediction is proposed. The robotic arm, which is made of multifilament muscles, can make similar movements to human arms, and its performance is verified in the simulation environment. We use the Kinect v2 depth camera and combine the depth image and color image to calculate the spatial coordinates of the target object. Finally, we use a neural network to predict the air pressure to control the robotic with a pressure controller. The experimental results show that the robotic arm can complete the spatial target tracking tasks with an average error of 2.14cm.