<p>U doktorskoj disertaciji rešavaju se sledeći problemi: problem opisivanja sistema masovnog opsluživanja kada teorija masovnog opsluživanja nailazi na ograničenja primene, problem predviđanja vremena čekanja, problem modelovanja odnosa na tržištu ekspres usluga kao izvora uticaja na redove čekanja, problem upravljanja brojem aktivnih kanala sistema masovnog opsluživanja i problem uticaja na subjektivno vreme čekanja. Primenom elemenata veštačke inteligencije i statističkih metoda razvijen je model za predviđanje parametra vremena čekanja u realnom vremenu pri jedinicama poštanske mreže za pružanje usluga korisnicima.</p> / <p>The dissertation provides answers to the following issues: the problem of describing the queueing system when the queueing theory encounters limitations in its use, predicting the waiting time, the problem of modeling relations in the market of express services as a source of influence on the queues, managing the number of active channels in the queueing systems and the impact on subjective waiting time. Through application of artificial intelligence and statistical methods, a model has been developed which in real time predicts the parameters of waiting time at the units of postal network that provide service to customers.</p>