Hybrid metaheuristic approach GA-SA for the buffer allocation problem that minimizes the work in process in open serial production lines

Abstract

The Buffer Allocation Problem (BAP) is a problem of combinatorial NP-Hard optimization in the design of production lines. This consists of defining the allocation of storage places (buffers) within a production line, in order to maximize the efficiency of the process. The methods of optimization have been reported with greater success in recent years are metaheuristic techniques. In this work, a hybrid approach is proposed that uses the metaheuristic techniques of Genetic Algorithms (GA) and Simulated Annealing (SA), with the objective of determining the required buffers that minimize the average work in process (WIP) in open serial production lines M/M/1/K. The evaluation is carried out with an analytical method of decomposition. The results obtained demonstrate the computational efficiency of the proposed hybrid algorithm with respect to a simple SA or GA.