Darbo tikslas yra išanalizuoti keliaujančio pirklio algoritmo realizacijos galimybes egzistuojančiose navigacinėse sistemose, bei išanalizavus pasirinktus algoritmus keliaujančio pirklio problemai spręsti, parinkti tinkamiausią algoritmą pagal turimus atminties ir skaičiavimo resursus bei problemos sudėtingumą. Tyrimo rezultatai parodė, jog nėra tinkamiausio algoritmo visiems atvejams, nes skirtingose situacijose skirtingi algoritmai rodo geriausius rezultatus. / The investigation had a list of objectives: analyze the capabilities and resources of a range of chosen GPS navigation devices; analyze the needs and requirements of traveling salesman related GPS navigator functions for regular users; analyze what types of TSP algorithms are used in existing navigation software products; analyze the capabilities of various TSP algorithms with regard to used resources and speed of calculations; determine which algorithms are optimal for a range of specific situations. Research of different algorithms led to a conclusion that there is no single algorithm that is always better than the rest. Under different circumstances, different algorithms showed different results. Some were clearly optimal in some situations, while others competed with each other in other situations. The key element to success of an algorithm was how much time it got to do it's calculations. The amount of the input data changed the duration of the calculations but the algorithm function declination rate remained mostly the same with different sets of input data.