Digital energy: as the intellectual analytics and the forecast of technical condition of the equipment transfer power to "digit"

Using the systems of the class AI — intellectual analytics — the energy companies can analyze all collected data and predict technical condition of the equipment. One of solutions for intellectual diagnostics and the predictive analysis is Clover PMM (Predictive Maintenance & Monitoring) — development of the Russian company Clover Group.

Solvable tasks using Clover PMM in power

At the heart of Clover PMM – two modules: search of anomalies (abnormal statuses) in operation of the equipment and the forecasting module which builds the forecast of failure of a structural element or the equipment.

The first module intellectually diagnoses the equipment – allows to reveal abnormal statuses in real time. For example, for one generation company Clover revealed exceeding of temperatures in turbine generator stator grooves. Exceeding was noncritical, but a system finds such cases and displays automatically. Secondly, a system understands that such situations will lead to a hardware failure, especially, if to continue to operate in the high-loaded modes. Therefore a system surely shows it to the user - it is and there is intellectual diagnostics.

The module of the predictive analysis predicts a status of units and studies in process of receipt of a feedback from engineers and technologists. This module allows to understand a status of a structural element during certain time before making the decision on repair or utilization of the equipment.

For example, according to the plan it is necessary to display the unit in repair in half a year. To this Clover already found exceeding of temperatures and now builds the forecast – what the anomaly whether it will be shown in the form of defect or failure when it happens – within half a year or later will lead to. So forecasting helps to make the decision on the fate of the equipment and to create the plan of repair on the basis of the actual and forecast state. Also Clover PMM doobuchatsya by engineers, technologists using machine learning.

The main objective which is solved by Clover PMM in power — the correct formation of the program of repairs of rather current and forecast technical condition of the equipment.

Users of Clover PMM

As Clover PMM is able to predict a stop, breakdown, to find prenegative statuses and displays this information of the computer, operation personnel which manages changes and the modes – the main users. For example, service of the chief engineer which is responsible for operation of the equipment. For them it is important to provide reliability of the equipment and uninterrupted operation of its work.

For repair and service companies Clover PMM it is convenient when planning material resources and structure of repair crews of power plant. Clover allows to see true and forecast technical condition on each unit as on the basic, and auxiliary, at the expense of it the program of repairs "on a status" forms.

Stages and implementation time of intellectual analytics

Analysts are the cornerstone these, main sources of which APCS ERP MES, the MRO, any other systems storing information from sensors about repairs; external factors (humidity, temperatures, etc.). Data are transferred to specialists Clover Group for preparation for the analysis: to structuring and enrichment.

It is important to understand that forecast accuracy depends on the number of data: if the customer saved up enough data, then the module of the predictive analysis will start working at once and it is possible to gain effects right now. If data are not enough, then after Clover PMM implementation charges only begin, and setup of predictive model will take several months.

Each equipment requires the different number of data. For example, for ancillary equipment, the compressor of Atlas Copco company was data of one year enough. After collecting data load into the expert and analytical Clover system where mathematicians together with industry experts select abnormal statuses and defects of operation of the equipment, identify violations of the modes of operation and build the forecast failure pattern of units.

A system doobuchatsya by industry experts and engineers, i.e. digitizes experience of experts and applies it in the analysis. For example, at breakdown of a structural element or emergence of anomaly, it is possible to classify the reason why so happened, for example, because of violation of the mode of operation. Knowledge of experts is for this purpose necessary, they explain what became the anomaly reason, and mathematicians write this rule in a system, doobuchy it on the basis of expert opinion.

For an output of results of Clover PMM it is integrated with the customer's IT infrastructure, for example, of the class ERP, EAM. Prenegative conditions of equipment, data on the actual and forecast technical condition of a structural element and the equipment are transferred to this system. If there is no management system for repairs, Clover is built in another — depends on features of infrastructure of the customer.

Demonstration of the Clover PMM interface for the blower

The term of implementation of the modules Clover PMM — 4 months.

It is important to understand that PMM is the system of analytics of Big Data which allows to pull out valuable grain in a type of physical dependences and on the basis of this information to make decisions on the equipment, but not classical industrial control system whose basic function – to manage the equipment on settings of parameters. These systems well supplement each other: An APCS gives data, and Clover PMM analyzes.

Effects and benefits for the energy companies

The greatest effect and benefits will receive the generation companies — those who directly face repairs and at whom the economy of the enterprise depends on operability of the equipment. For example, Clover PMM will help to avoid penalties because of abnormal shutdown of the unit.

Short-term benefits it will be visible from the first days of work — optimization of planning regarding the order of necessary spare parts and repair materials of the equipment, reduction of number of unscheduled repairs, cutting of costs for repair of the equipment, control of violations of the modes of operation.

Key long-term effect — change of repair regulations, repair on a status, and this reliability augmentation of the equipment and its "understanding".

190