Optimal robot task scheduling based on adaptive neuro-fuzzy system and genetic algorithms
- PDF / 1,840,132 Bytes
- 13 Pages / 595.276 x 790.866 pts Page_size
- 83 Downloads / 189 Views
ORIGINAL ARTICLE
Optimal robot task scheduling based on adaptive neuro-fuzzy system and genetic algorithms E. Xidias 1 & V. Moulianitis 1 & P. Azariadis 1 Received: 19 August 2020 / Accepted: 24 September 2020 # Springer-Verlag London Ltd., part of Springer Nature 2020
Abstract Industrial manipulators should be able to execute difficult tasks in the minimum cycle time in order to increase performance in a robotic work cell. This paper is focused on determining the near optimum route of a manipulator’s end-effector which is requested to reach a predefined set of demand points in a robotic work cell. Two subproblems are related with this goal: the motion planning problem and the task scheduling problem. A new approach is presented in this paper for simultaneously planning collision-free motion and scheduling time near optimum route along the demand points. A combination of a geometrical approach and an adaptive neuro-fuzzy system is employed to consider the multiple manipulator’s configurations, while a special genetic algorithm is designed to solve the derived optimization problem. The experiments show that the proposed method has the capacity to determine both the near optimum manipulator configurations and the near optimum sequence of demand points. Keywords Task scheduling . Manipulator . Adaptive neuro-fuzzy system . Genetic algorithms . Collision avoidance . Robotic work cell
1 Introduction Nowadays, robotics and more specifically industrial manipulators play an important role in Industry 4.0. Industrial manipulators have several features such as flexibility and adaptability which make them attractive for various assignments within large industrial environments. Moreover, they must be able to perform their tasks as fast as possible in order to keep productivity in a high level and simultaneously to decrease production costs [1, 2]. Thus, one of the major topics of the robotics research community is to develop methodologies that enable the manipulators to perform the requested tasks, such as spot welding, as
* E. Xidias [email protected] V. Moulianitis [email protected] P. Azariadis [email protected] 1
Department of Product & Systems Design Engineering, University of the Aegean, Ermoupolis, Greece
quickly as possible, considering the limits imposed by (a) their characteristics and (b) the working environment [3]. The presented work is motivated by industrial applications where a manipulator is requested to visit a set of demand points with no predefined order. Some examples of such applications are the laser cutting, the multiple drilling and the spot welding [4]. There are numerous industrial activities where the manipulator is requested to reach several demand points and return to the initial demand point. Clearly, in these problems, the sequence of demand points has a strong effect on the total cycle time. The problem of optimal multi-tasking motion planning for an industrial manipulator which is requested to operate in a robotic work cell can be considered the coupling of two NPhard problems: (a) the m
Data Loading...
