Design and built rubik machine applying deep learning technology

Abstract

Research, design, and manufacture a robot that rotates the Rubik according to the model: 6 3x3 block clamping mechanisms (side size 57mm), the robot frame has 6 joints to rotate the corresponding faces of the Rubik's cube. Design and construction of a control circuit for a robot to rotate Rubik with the following functions: Position control of 6 stepper motors. Research on Rubik's rotation algorithm. Programming the Rubik rotation algorithm according to the mechanical model of 6 rotating joints. Research application of image collection and processing to identify Rubik's cube. Programming the console to solve Rubik's cube on a computer using AI deep learning. Achievements: Finished construction of the mechanical model according to the design, the control circuit works well and stably. The algorithm program completes the average Rubik rotation problem in about 30 steps. The image processing program recognizes colors accurately in well-lit environments