This work is licensed under a Creative Commons Attribution-NonCommercial 4.0 International License
|
||||||||
|
Paper Details
Paper Title
Performance improvement of power generation in gas turbines by using random forest and recurrent neural networks
Authors
  Mrs. T. Sujatha,  Mrs. S. Jayasree,  Mr. M. Kannan
Abstract
In order to improve the power generating performance of gas turbines, machine learning based models are developed. In this paper, random forest and recurrent neural network algorithms are combined to improve the power generation. Average errors based on the operating characteristics are captured by this model. Furthermore, these models are established for capturing the gas turbine part-load performance and full-load performance. Simulation performance is computed by embedding the air compressor and air turbine and individual dataset are used for the validation of these models. The correction curves of gas turbine performance are constructed by predicting the full load performance of RNN model and it possesses reduced complexity and increased accuracy. The obtained curves of the RNN model to predict the part-load performance produces extreme results for continuous turbine monitoring and diagnosis of fault. This proposed method is suitable to any gas turbines and it will be aided to all the power plants for studying the quantitative performance degradation with respect to time.
Keywords- RNN, air turbine, random forest, recurrent neural networks, gas turbine, complexity.
Publication Details
Unique Identification Number - IJEDR2204006Page Number(s) - 46-49Pubished in - Volume 10 | Issue 4 | December 2022DOI (Digital Object Identifier) -    Publisher - IJEDR (ISSN - 2321-9939)
Cite this Article
  Mrs. T. Sujatha,  Mrs. S. Jayasree,  Mr. M. Kannan,   "Performance improvement of power generation in gas turbines by using random forest and recurrent neural networks", International Journal of Engineering Development and Research (IJEDR), ISSN:2321-9939, Volume.10, Issue 4, pp.46-49, December 2022, Available at :http://www.ijedr.org/papers/IJEDR2204006.pdf
Article Preview
|
|
||||||
|