Integration of results of monitoring technical conditions of transfer pipelines

The existing systems of monitoring of transfer pipelines include control over the technical condition of linear part and equipment; assessment of technological processes impact on the environment; analysis of qualitative indicators of material flows at the entrance and exit of equipment [1].
The main issues of users are limited access to the data due to isolation of monitoring systems from each other and absence of specially developed algorithms of joint processing of data of a diffirent physical nature. To solve such issues, the present article provides an analysis of existing systems of monitoring of transfer pipelines, defines main controlling parameters, suggest an approach to integration of monitoring systems of architecture and data. The practical value of this article is adaptation of such solutions as Big Data and Data Mining methods while establishing data centres at enterprises of oil and gas industry to integrate, synchronize and analyse the technical information.
Materials and methods
Projections of software, intellectual data processing.
Results
It was defined the similarity of architectures of monitoring systems, that is characterised at the lower level by transducer of physical processes and intermediary centres of datacollection, and at the medium and upper levels by local and global centres of data processing, that help to analyse the technical conditions of an object. At the bottom of the supposed system there are data flows and methods of retrieving the knowledge, which enable to form managerial impact by combining such knowledge with information archives already in place [21].
Using the algorithms of machine learning, identifying consistent patterns in the collected data, it is possible to conduct assessment and forecast of the technical conditions of transfer pipelines.
Сonclusions
Integration of data received from different transducers and measuring equipment radically increases the efficiency of monitoring technical conditions of a transfer pipeline. Since the volume of received information is exponentially growing with time, and significant part of this information is rarely required, it is presented an approach of cheapening of storage systems based on NoSQL databases. The suggested solution enables to work with structured and unstructured information at the same time.