Solving the energy-efficient Robotic Mixed-Model Assembly Line balancing problem using a Memory-Based Cuckoo Search Algorithm

Minimizing energy consumption is an important issue in robotic assembly lines where a set of robots are assigned to a set of workstations to perform different tasks. When it is planned to assemble several models of one product in the same robotic assembly line, the minimization of energy consumption becomes more difficult since the best assignment of tasks and robots to workstations must be found, taking into consideration all models. The authors cannot find in the literature a work that aims to minimize energy consumption in robotic assembly lines that produce several models with one configuration. Furthermore, the introduction of the heterogeneity of models and robots makes the problem more complex and hard, even for small-scale instances, and for this reason, the authors propose in this paper a Memory-Based Cuckoo Search Algorithm (MBCSA) to tackle this problem. The principle of memory is used in this new Cuckoo Search Algorithm in order to escape from the local optima and discover new search zones. Six problems of different sizes are generated and solved by the proposed MBCSA, and to evaluate its performance, two comparisons are made with two meta-heuristics, the genetic algorithm and another version of the cuckoo search algorithm. Obtained results show that this new version of the Cuckoo search algorithm is promising and can obtain good solutions for problems of different sizes.