The paper describes an approach for determining the technical state of an axial compressor as part of a gas turbine unit, based on thermogasdynamic parameters that are measured at the operating facility. As a criterion for the technical condition, it is proposed to use the ratio of the actual efficiency of an axial compressor to the reference efficiency in such modes. To do this, it is enough to know the temperatures and pressures of the air at the inlet and outlet of the axial compressor. A method for selecting such modes is described and some features of filtering the operating parameters of a gas turbine plant are noted. In the proposed approach, the choice of modes for analyzing the technical state of an axial compressor is carried out according to the value of the effective power of the gas turbine unit. Some results of determining the criterion of the technical condition of an axial compressor for a gas turbine unit, which drives a centrifugal compressor of natural gas, are presented. Based on the analysis of the degradation of the technical condition criterion, it is possible to determine its predicted value. The results of the work can be used to create automated systems for assessing and predicting the technical condition of drive gas turbine units at their facilities. The application of the obtained results of the work expands the possibilities of servicing gas turbine plants according to the actual state. The proposed approach can be adapted and applied for gas turbine plants for various purposes, as well as for their individual units.
This paper considers the method to estimate the technical condition of gas turbine power for natural gas transportation, using machine learning methods. Source data was used to archive gas-dynamic parameters from the automatic control system of the gas turbine. The method is based on changing the enthalpy of the natural gas before and after the centrifugal gas compressor is used for creating a dataset with measured parameters and power from the gas turbine. The actual power is determined from the line of modes for a certain period. The software is implemented using Python and the Scikit-learn library is used to create machine learning models. A mean average percentile error is chosen as the model quality criterion. In this paper, different sets of feature parameters and sample sizes are researched by the quality of the prediction machine learning models. Recommendations on the use of models are given. It has been established that the approach is not applicable for predicting future technical condition without the presence of data on a similar technical condition in the training sample. It is recommended to use the described approach to determine the technical condition in a period of operation in the past.
Erosive wear of the parts of the gas path of an axial compressor of a gas turbine is a common reason for premature decommissioning of equipment. The creation of an advanced diagnostic system, which will allow determining the level of blade erosion according to standard parameters without the inspection or disassembly, is topical for Russian gas transmission enterprises. The paper presents preliminary results of applying machine learning methods to solve such a problem for an isolated stage of an axial compressor. The verified results of numerical simulation of the air flow in the stage were used as initial data. The degree of erosion was set as the ratio of the chord of the eroded blade to the chord of the new blade in the peripheral section. The same parameter was the target for machine learning models. Sets of local and integral parameters of the numerical calculation were used as parameters. As a result of the primary study, the random forest model showed the best results when using all available parameters and the parameters with the highest correlation. Conclusions are formulated about the applicability of machine learning methods for creating a model for assessing the degree of erosion. The development of the work is connected with the creation of a model for predicting the technical condition of the flow path of the entire compressor.
The paper considers methods of the gas turbine plant power designed for natural gas transportation and reveals their drawbacks. A program in the Python language was created to study applicability of the machine-learning methods to determine the plant power under operating conditions. Archival gas-dynamic parameters registered by the plant automatic control system were used as the initial data. Forecast quality of the machine-learning models was estimated depending on different sets of the feature parameters. Recommendations on the models use are provided; and the method error was determined. Hypothesis on applicability of models learned based on data of a single engine to estimate the power of the other engines of the same type was refuted. Machine-learning methods could be used to determine the gas turbine plant power even in the absence of part of the initial data, which is the main advantage over traditional methods.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations –citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.
hi@scite.ai
10624 S. Eastern Ave., Ste. A-614
Henderson, NV 89052, USA
Copyright © 2024 scite LLC. All rights reserved.
Made with 💙 for researchers
Part of the Research Solutions Family.