INTEGRATION OF ROBUST SHORTEST PATH WITH PICKUP AND DELIVERY VEHICLE ROUTING PROBLEM

The high changeability of vehicle speed in the urban network increases the level of uncertainty as to the estimated travel times. The rush hours, location of road sections, weather conditions or traffic incidents are just some of the factors which influence vehicle speeds. These and other factors are accountable for a significant difference between the estimated speed and the actual value. As a consequence, the designated routes of vehicles based on incorrect and obsolete data may be non-optimal or even infeasible. Such situation may lead to an increase in transportation costs borne by the delivery company operating in an urban area. Moreover, non-optimal routes could create augment costs in the entire transportation system e.g. due to noise and pollutant emission, increased road occupancy or damage to the touristic image of the city. The article presents an approach based on integration of the shortest path problem with the pickup and delivery vehicle routing problem. Through the integration of both methods it is possible to obtain exact solutions to robust vehicle routing problems within a reasonable time. The analysis and the characteristics of the robust model are presented on a real example - the road network of Cracow. Data and characteristics of traffic were obtained from road traffic simulator Aimsun 8.0 and the optimization was performed using CPLEX 12.5 environment.