Optimal buffer allocation and service rates in flow line production system

In this research, we present a method to solve for an optimal solution vector containing buffer allocation and service rates of the flow line production (FLP) system such that the throughput is maximized. The solution method integrates the elitist teaching-learning-based optimization (ETLBO) and optimal computing budget allocation (OCBA). At first, The FLP system is formulated as an integer-valued inequality constrained optimization problem with a large search space. The ETLBO is then utilized to select N excellent solutions from the search space, where the objective value is evaluated with the radial basis function (RBF). The RBF is taken as a meta-model to approximately estimate the objective value of a solution vector. Lastly, the OCBA scheme is adopted to look for an optimal solution vector. The solution method is tested on two examples of FLP system, one comprising 3-stage and another comprising 12-stage. Simulation results present that the superiority of the solution method in the solution quality and computing efficiency using extensive simulations.