SOLVING SMALL TSP ACCORDING TO THE PRINCIPLE OF MINIMUM ACTION

To solve the well-known Travelling Salesman Problem (TSP), many solutions based on combinatorial optimization, heuristic and meta-heuristic have been proposed. However, in managing business processes, a few times we attend to the real time optimization of picking routes either inside a warehouse or within materials or waste recovery distribution systems. This study proposes a new algorithm which is based on the analogy between TSP and conduction heat transfer; in particular, the application of the principle of minimum action to the heat transfer of a flat plate, which is coincident with the physical domain, over which the TSP points stress, helps identifying the order sought. The algorithm has been implemented in an Excel® spreadsheet; the quality of solutions which have been found is midway between the nearest neighbor algorithm and a genetic one; data processing time appears suitable for logistic processes management.