Algorithms for estimating throughput characteristics in a generalized call center model

Resumo

We continue the study of a generalized call center model presented in [1]. The model takes into account the following characteristic features of reference and information services: division of the servicing personnel into operators and consultants; the possibility of repeating a claim when operators, consultants, or access lines are busy, or due to a failed end of waiting time; the presence of automatic responding machines. For the considered model, we construct approximate algorithms for estimating stationary characteristics of servicing incoming claims. The algorithms are based on using special cases of the model in question, on applying asymptotic decompositions of characteristics as the repetition intensity tends to zero, and on implementing the model decomposition principle into individual segments with specially tuned values of input parameters. We show numerical examples that illustrate the errors of the resulting computational procedures.