Neuro-Fuzzy Intelligent Control Algorithm for Cable-Driven Robots with Elastic Cables

Abstract

First-generation industrial robots were made of solid manipulators which were connected by several consecutive joints that are known as serial mechanical manipulators. The need for robots has increased in time with modern industrial development because of the need for high accuracy, rapid acceleration and the cargo-carrying abilities of parallel robots. Parallel robots are difficult to implement in limited workspaces, which led to the development of parallel robots connected by cables instead of solid manipulators. This new structure uses cable-driven tensional force and can be used to design parallel robots with a large workspace. In terms of weight, a cable-driven robot is lighter than a common robot. The use of a cable in the structure of the robot to apply tensional force has created new challenges. The control algorithms used in serial and parallel robots had not been designed for cable-driven robots; thus, new algorithms were required to meet the needs of such robots. The current study used simulation to model a parallel planar cable-driven robot with three degrees of freedom using an adaptive network-based fuzzy inference system (ANFIS). In addition to stabilizing this novel planar cable-driven robot, the ANFIS control algorithm allows better detection of reference routes.