Research On Installation, Modeling And Control Of The Constrained Flexible-Link Robot Manipulators

In recent years, more attention has been paid on the study of constrained flexible-link robot manipulators. The research has great significance to enlarge robot applications in industry.After summarizing the study of constrained flexible-link robot manipulators, this dissertation proposes the program to manufacture the flexible-link robot manipulators. Based on this program, One constrained flexible-link robot manipulators is designed, machined and installed by author himself. Chapter 3, studies the drive and protection of the manufactured robot manipulators and in other chapter, studies the key techniques of manufacturing the flexible robot manipulators, such as data communication, signal testing and disposing.Based on the D'Alembert-Lagrange principle, Chapter 4 discusses the dynamic modeling problem for a class of constrained planar two-link flexible manipulators, and establishes a set of dynamic equations which describe the motion of the robotic systems of our manufacturing robot. Compared with the present dynamic models for same robotic manipulator systems. the motion equations and vibration equations built up in the dissertation are more precise and simpler. Moreover, this dynamic model has a similar configuration with unconstrained rigid-link manipulators, which make it easy to control the complicated constrained flexible manipulators.Chapter 5 and Chapter 6 deal with the position control of the constrained flexible robot. Chapter 5 proposes and designs a direct adaptive controller on neural network. This controller used the neural network to identify the model of flexible. Based on the PID control , Fuzzy control and the knowledge of expert, Chapter 6 proposes and designs an expert PID controller, the parameters of PID are modified on line by the upper computer. The result of control verified on the constrained flexible robot shows that these two kinds of controller can control the mechanical arm to position accurately. Chapter 7 and Chapter 8 deal with the position/force control of the constrained flexible robot. Chapter 7 simplifies the dynamic model established in Chapter 4 reasonably and computes the law of control by the computer. According the law of control, controls the flexible-link manipulators to move on the constrained trajectory and hold the force on the constrained surface to be constant. The result of control is good. In order to get better, Chapter 8 proposes a hybrid position/force control on neural network method. This controller used neural network to compensate the uncertain effect on the flexible robot. The simulation result of this controller shows that this method is effective.