Optimal regulation of a cable robot in presence of obstacle using optimal adaptive feedback linearization approach

Abstract

Optimal path planning of a closed loop cable robot, between two predefined points in presence of obstacles is the goal of this paper. This target is met by proposing a new method of optimal regulation for non linear systems while Dynamic Load Carrying Capacity (DLCC) of the robot is supposed as the related cost function. Feedback linearization is used to linearize the system while Linear Quadratic Regulator (LQR) is employed to optimize the DLCC of the system based on torque and error constraints. Obstacle avoidance for both the end-effector and cables is also considered by the aid of designing an adaptive local obstacle avoidance controller. As a result of linearized nature of the proposed optimal regulation and obstacle avoidance, fast calculation for real time applications is possible. Therefore, formulation of the optimal feedback linearization, together with calculating the DLCC of the robot based on the presented constraints is derived. Finally, a simulation study is performed to study the optimal dynamics and also the maximum DLCC of the cable robot in presence of obstacles. Simulation results are eventually compared with experimental tests conducted on IUST Cable Suspended Robot (ICaSbot) to verify the validity and efficiency of the proposed optimal controllers.

On the Active and Passive Flow Separation Control Techniques over Airfoils

“…Controlling the behavior of mechanical systems is an integral concept that inevitably arises in entire engineering research fields from dynamics and robotics [9][10][11] to fluid mechanics [12][13][14] and turbulent boundary layers [4][5][6][7]. Boundary layer can be controlled through a passive method, which not needs a controller, and/or an active method, which requires a controller and controlling sensors [8].…”

SDRE controller for motion design of cable-suspended robot with uncertainties and moving obstacles

“…For cable-suspended robots, many control methods have been proposed; openloop controllers like Hamilton-Jacobi-bellman (HJB) [10] (which is weak against noises and uncertainties), closed-loop controllers like feedback linearization [11] and Sliding Mode Control (SMC) [12] (which are not optimal), Optimal linear and closed-loop controllers like Linear Quadratic Regulator (LQR) with feedback linearization basis [13] (which is unable to capture nonlinearities properly).…”

“…Motion design for cable-suspended robots that perform in huge workspace might be problematic and are usually subjected to different constraints like obstacles [13] and moving boundaries [18]. In this paper, obstacles were considered based on their more commonalities over other constraints.…”

“…Most of the existing obstacle avoidance techniques are based on the APF concept [19]. The potential term created by this method is usually nonlinear which is problematic for linear based control methods like LQR and makes them more complex and restricted [13]. While SDRE is shown to be capable of being combined with this technique very conveniently and perform efficiently in the presence of fixed obstacles [14], many other potentials of such controllers are not yet studied.…”