There are two general methods to control cooperative mechanical manipulators: Centralized & Decentralized. In decentralized method, each manipulator is controlled separately and each manipulator’s controller doesn’t know the other one’s situation. The effect of one to the other manipulator is considered as disturbance. In centralized method the set of cooperative manipulators is considered as a unified system and only one controller is designed for the whole system.
The cooperative manipulators make a closed chain system with redundant actuators. Actuator redundancy makes it possible to optimize a dynamic index and control of the system simultaneously. If centralized controller is used, there are two methods to design the controller. First: turning off the redundant actuator and designing a suitable controller. Second: using the redundant actuator as a merit and designing a controller that controls the system and optimizes a dynamic index simultaneously.
In this project after formulating the dynamical equations of the cooperative manipulators in cases of carrying a point mass and defined length and mass object, the centralized & decentralized controlling methods have been designed and tested numerically.
For systems with less complexity a real-time exact optimization and for systems with more complexity a real-time near exact optimization has been proposed and tested numerically.
In this project, to verify the proposed methods experimentally, two cooperative planner dual link manipulators have been designed and made. A summery of the designing, making and setting up the system procedure has been given.
At the end numerical simulations have been tested experimentally and results have been shown.