V. R. Milov – Dr.Sc. (Eng.), Professor, Head of Department «Electronics and Computer Networks», Nizhniy Novgorod State Technical University n.a. R.E. Alekseev. E-mail: email@example.com
E. L. Karnavsky – JSC «Giprogascentе». E-mail: firstname.lastname@example.org
S. A. Nikulin – Engineer, JSC «Giprogascentе». E-mail: email@example.com
R. L. Shibert – Ph.D. (Eng.), Associate Professor, Nizhniy Novgorod State Technical University n.a. R.E. Alekseev. E-mail: firstname.lastname@example.org
One way to reduce corrosion of trunk pipelines is the electrochemical protection (ECP), in particular, the cathodic polarization, which provides cathodic protection stations (CPS). With the introduction of remote corrosion monitoring built using telemetric controlling and measuring points, there is a possibility of realization of auto-control means of ECP examined by determining the optimal functioning of CPS based on the processing of the recorded data. The problem of identification of the «pipe–ground», which consists in the formation of dependence of the voltage at the control point of the vector currents cathodic protection stations on the basis set of multiple linear regression models. Special attention is given to selection of significant input variables (covariates). Apriori information on the relative positions of control points and controlling and measuring points allowed to obtain a non-stepwise regression search algorithms based on sequential exception of input variables of the full model. An alternative methodically integral approach to the problem of structural-parametric identification is the Bayesian methodology. The results obtained in the framework of this methodology algorythms structural-parametric synthesis of neural network models obviously are reduced to the case under consideration for determining the structure of the linear multiple regression model. The procedures of model verification and analysis of the reliability of monitoring data use a priori information about the values and the ratio of the parameters of regression models are considered. Field experiments conducted at the site of the main gas pipeline «Saratov–Gorky» confirm efficiency of the procedures for the identification of the «pipe–ground», which determines the possibility of an operational definition of the optimal functioning of CPS. The approach developed in view of solving the problems of the adaptive tracking seasonal changes and processes of degradation of the insulation coating, as well as the prediction of corrosion condition plots of trunk pipelines is the basis of mathematical software to develop automated control systems by means of ECP. The introduction of the proposed system will facilitate the timely generation of rational decisions operational management system of electrochemical protection, its maintenance and repair.