Properties and structure of dynamic robot models for control engineering applications

Controller design for robotic manipulators requires a fundamental physical understanding of the properties and structure of dynamic robot models. This paper focuses on the Lagrangian formulation which is attractive from both the dynamic modeling and control engineering points-of-view. Physical and mathematical properties and structural characteristics of the complete dynamic robot model are demonstrated. Implications of the model for control system analysis and design are then indicated. Physical interpretation leads naturally to the decomposition of the model into the positioning arm and end-effector subsystems and motivates the application of decentralized control to robotic manipulators. The authors then propose the application of control the positioning arm and artificial intelligence and intelligent sensors to control the end-effector.