In rail freight transport, the primary carrier, PKP Cargo S.A., achieves very good results in full-train shipments, including international transport. Meanwhile, freight transport in single wagons or small groups of wagons has decreased overall from approximately 80% in 1980 to about 30% today.
The focus of PKP Cargo S.A. on full-train shipments, which is reprehensible from the perspective of rational transport service for the country, stems from the fact that they are economically efficient and technologically simple, whereas wagonload shipments are less economically efficient, more labor-intensive, and quite technologically complex. Wagonload shipments in their current form, due to their questionable attractiveness, play a smaller role in the country’s transport service than is justified and possible. The need to rationalize transport processes to counteract the marginalization of rail transport in this market segment is obvious. Under market conditions, with the dominant role of road transport, the demand for wagonload shipments will exist when the system for organizing wagon flows is optimal. Meanwhile, the methods used for this organization to date are practically heuristic and do not guarantee optimal solutions. This refers to the classical method of A.P. Petrov and even the computer method of S.W. Duwalian, which is completely unknown in Poland.
Both of these methods do not take into account constraints related to the number of directional tracks and the station’s handling capacity. Furthermore, Duwalian’s computer method requires very labor-intensive manual operational correction. This paper presents a branch and bound method that eliminates the drawbacks of the methods mentioned above. The idea of this method is to utilize the finiteness of the set of variants (feasible solutions) and to move from a full to a limited review of them. An important role is played by the evaluation and rejection of unpromising (non-optimal) subsets of variants. Unlike methods that sequentially combine the most profitable relations, the branch and bound method builds a structural interdependence of relations. This guarantees obtaining an optimal train formation plan with limited resources (directional tracks or station handling capacity).
The article contains 22,000 characters.
Source: Logistyka Magazine