Optimización de manipulabilidad y consumo eléctrico mediante el Algoritmo Heurístico de Kalman en manipuladores seriales

Abstract

En el presente trabajo se plantea una metodología de planeación de trayectorias para manipuladores seriales antropomórficos de seis grados de libertad y muñeca esférica enfocada en la minimización del consumo eléctrico y maximización de la manipulabilidad. Para lograr tal fin se expone un algoritmo de optimización el cual tiene como base el Algoritmo Heurístico de Kalman (AHK), cuya finalidad es encontrar la trayectoria óptima según la función multi-objetivo propuesta dentro de un espacio esférico simplificado propuesto. El algoritmo de optimización se evalúa mediante el análisis del comportamiento de dos trayectorias en el manipulador PUMA 560, simuladas en un programa desarrollado en la aplicación de Visual Basic perteneciente al software Autodesk Inventor.

Energetically Optimal Trajectory for a Redundant Planar Robot by Means of a Nested Loop Algorithm

“…Promising results have been obtained at the simulation level and in real implementations. This is the case of the technique of the subspace trust method [2], Euler/Runge Kutta with multiple method shooting [3], dynamic programming with Dubins routes [4], dynamic programming "Mixed Integer Nonlinear Programming" [5], Interior Point Optimizer (IP) [6], evolutionary model and swarm intelligence [7], differential evolution [8], gradientbased algorithm [9], dynamic programming [10], Kalman heuristic algorithm [11], Genetic Algorithm (GA) [12], metaheuristic algorithm type vector evaluated particle swarm optimization [13], sequential quadratic programming [14], Hessian and GA matrix [15], Pontryagin's minimum algorithm [16], and optimization of multiple immune targets of restrictions [17]. Finally, in [18], a trajectory planning method is developed based on a surrogate or substitute model for an unmanned electric excavator.…”

“…Similarly, to determine the energy used by the robotic manipulator, using an objective function, it is possible to consider different variables obtained from the RE operation Energetically Optimal Trajectory for a Redundant Planar Robot by Means of a Nested Loop Algorithm such as total work [13], some squared torque variants [3], [14], [19]; squared current [12], actuator motor power [5], [11], [15], [16]; squared acceleration [4], [17], [20]; potential and kinetic energy [8], mean square torque [9], [21] and mechanical power [9]. In addition, the authors consider the use of a weight vector that penalizes the influence of each joint in the optimization task, such as Zhao, Lin, and Tomizuka [22] and Wigstrom, Lennartson, Vergnano, and Breitholtz [5].…”

“…In [12], a polynomial formulation of the 3 rd and 4 th degree is used. Finally, models such as the Rubins route and predesigned models are found to be advantageous in [4] and [11], respectively.…”

“…Subsequently, from the equations developed from the homogeneous matrix (11) and suitably solving for 2 3 , T it is possible to determine θ 1 , obtaining the expression (18) 12 .…”