Job shop scheduling problem based on learning effects, flexible maintenance activities and transportation times

Nowadays, scholars do their best to study more practical aspects of classical problems. Job shop Scheduling Problem (JSSP) is an important and interesting problem in scheduling literature which has been studied from different aspects so far. Considering assumptions like learning effects, flexible maintenance activities and transportation times can make this problem more close to the real life, however these assumptions have rarely been studied in this problem. This paper aims to provide a mathematical model of JSSP which covers these assumptions. MILP model is suggested, Three different sizes of instances are generated randomly, and this model has been solved for small-sized problems exactly by GAMS software and the effects of learning on reducing the value of objective function is shown. Due to the complexity of the problem, in order to obtain near optimal solutions, medium and large instances are solve by applying Ant Colony Optimization for continuous domains(ACOR) and Invasive weed Optimization (IWO) algorithm, finally results are compared.