A two-phase Pareto front method for solving the bi-objective personnel task rescheduling problem

This paper considers the personnel task rescheduling problem, which arises due to the occurrence of operational uncertainty, making the initial baseline schedule infeasible. To restore the feasibility, the personnel task rescheduling problem should be solved, which considers two conflicting objectives, i.e. the maximisation of the service level and the minimisation of the resource schedule deviations. To facilitate the decision making process and decrease the subjectivity of the decision-maker, a multi-objective optimisation approach is designated to offer a set of alternative solutions to the different stakeholders, making an adequate trade-off between both objectives to improve schedule acceptance. To that purpose, we present a two-stage multi objective decomposition approach targeted at generating a set of evenly spaced non-dominated solutions that compose a representative Pareto front. The first stage comprises an adaptive weight-based method to construct an initial set of solutions. The second stage of our approach further explores the solution space by inducing a constraint-based method to improve the distribution of the solution points. Several computational experiments are conducted to evaluate the solution quality and computational efficiency of the proposed method. We benchmark the performance of the two-stage approach with other multi-objective solution approaches constructing a Pareto front and validate the design choices made in both stages.