In the presented article, problems associated with the modeling features in automation and control of processes occurring in a primary oil refining technological installation are studied and investigated. In the article, when solving the modeling problem, special attention is paid to the problem of increasing the completeness of the received information used for processing, and in this connection, based on an active experiment, the “noise” characteristics inherent to the main indicators of the technological process under study are determined. It also addresses issues related to the correlation regression analysis features and the basic principles of designing an optimal filter that ensures the necessary cleaning of the received information from harmful interference and noise inherent in information communication lines. The article proposes an adaptive filter with an adaptation algorithm, which, analyzing the current state of the process under study and taking into account the disturbing influences constantly acting on the technological process, promptly corrects the optimal tuning parameters of the filter, minimizing distortion of the useful signal for the automatic control system.