A Hybrid Resource Allocation for Robust Scheduling ThroughCombination of Neural Networks, Critical Path Method andNSGA-II

In many scheduling problems with controllable processing times,commonly two general resource consumption function is used, linear andconvex, and the effect of the starting time of activity not considered.However, both of these functions have some assumption fails such as theeffect of the environmental condition from starting time of activity andquality of assigned resources. Inspired by these perceptions, we providea hybrid model for robust project scheduling in which each activity’sprocessing time is a function of its starting time and quality and quantityof its allocated resource. The objective is optimizing sequence ofactivities respect to time and cost of project with considering limits ofavailable resources. To achieve this goal, in the proposed method, NeuralNetworks is combined with Critical Path Method for calculation ofactivity’s processing time, then NSGA-II is used for optimization. Forevaluation of model capability, a problem is optimized with proposedmethod. The objective function is to minimize the makespan and totalcompletion cost of project. The results show that the model has factorsof robust scheduling

Keywords:

Scheduling , Resource Allocation , Artificial Neural Networks , NSGA