A constraint-programming based decomposition method for the Generalised Workforce Scheduling and Routing Problem (GWSRP)

Abstract

This paper deals with the Generalised Workforce Scheduling and Routing Problem (GWSRP) where 9 temporal constraints ensuring visit dependencies are all together taken into account and where customer and worker’s quality of service are taken into consideration. A Constraint-Programming based Decomposition Method (CPDM) is proposed, firstly based on a relaxation of coordination constraints and a column generation, and secondly with an iterative insertion of coordination constraint by constraint programming solver. Numerical experiments are achieved on huge instances derived from WSRP benchmark instances with up to 177 customers, 59 vehicles and coordination constraints. The CPDM is able to find nearly optimal solution for medium-size instances and find high-quality solution for huge-size instances whereas CPLEX solver applied to a mixed integer linear model is not able to give a solution in this case.