Development of a Dynamic Neural Network Model for Multistep ahead Prediction of Exhaust Gas Temperature in Heavy Duty Gas Turbines (original) (raw)
Related papers
A gas turbine thermal performance prediction method based on dynamic neural network
IOP Conference Series: Materials Science and Engineering, 2021
In order to ensure safety and reliability of energy transportation, it is necessary to understand and predict the performance of the gas turbine components. A prediction frame of the gas turbine compressor isentropic efficiency is established using the neural time series theory based on the Dynamic Neural Network. In order to obtain appropriate parameters for the network, a validation set is introduced to generalize the model. The compressor isentropic efficiency can be predicted based on the suggested model which provides an effective technical mean for the early warning of gas turbine performance. The experiment verified that the performance calculation model and the isentropic entropy efficiency prediction model based on the neural time series are effective.
Estimating Gas Turbine Internal Cycle Parameters Using a Neural Network
Volume 5: Manufacturing Materials and Metallurgy; Ceramics; Structures and Dynamics; Controls, Diagnostics and Instrumentation; Education; General, 1996
We show that a neural network can be successfully used in place of an actual model to estimate key unmeasured parameters in a gas turbine. As an example we study the combustion reference temperature, a parameter that is currently estimated via a nonlinear model inside the controller and is used in a number of critical mode-setting functions within the controller such as calculating the fuel-split between various manifolds. We show that a feedforward neural network using simple back propagation learning can be built for estimating combustion reference temperature. The neural network matches the accuracy of the current estimate; and it is more robust to errors in its internal parameters. This is advantageous from the point of view of implementation since a number of errors creep in when running the algorithm on a digital controller, and an estimator that is not robust with respect to such errors can degrade the performance of the whole system.
IAEME PUBLICATION, 2014
The use of Neural Networks (NN) has been proved to be a cost-effective technique. It is very important to choose a suitable back propagation (BP) algorithm for training a neural network. While these algorithms prove to be very effective and robust in training many types of non-linear multivariable modeling, they suffer from certain disadvantages such as easy entrapment and very slow convergence. The power generating industry is undergoing an unprecedented reform. NN is applied to predict coal properties, economic load dispatch, emission prediction, temperature control, etc. This paper compares the performance of the six neural network methods to predict nitrogen oxides emission from a 500 MW coal fired thermal power plant. The various BP training algorithms used are gradient descent, gradient descent with momentum, variable learning rate with momentum, conjugate gradient back propagation, Quasi-Newton BFGS Algorithm and also LevenbergMarquardt. The parametric field experiment data were used to build neural network. The coal combustion parameters were used as inputs and nitrogen oxides as output of the model. The predicted values of the model for full load condition were verified with the actual values.
Sustainability, 2022
Accurate simulations of gas turbines’ dynamic performance are essential for improvements in their practical performance and advancements in sustainable energy production. This paper presents models with extremely accurate simulations for a real dual-fuel gas turbine using two state-of-the-art techniques of neural networks: the dynamic neural network and deep neural network. The dynamic neural network has been realized via a nonlinear autoregressive network with exogenous inputs (NARX) artificial neural network (ANN), and the deep neural network has been based on a convolutional neural network (CNN). The outputs selected for simulations are: the output power, the exhausted temperature and the turbine speed or system frequency, whereas the inputs are the natural gas (NG) control valve, the pilot gas control valve and the compressor variables. The data-sets have been prepared in three essential formats for the training and validation of the networks: normalized data, standardized data ...
Nonlinear Gas Turbine Modelling : A Comparison of Narmax and Neural Network Approaches
2002
In this paper two nonlinear modelling approaches are employed to derive single nonlinear models for a Rolls Royce aircraft gas turbine. The first approach is based on the estimation of a NARMAX model using conventional structure selection and parameter estimation techniques, and the second approach is based on the use of feedforward Multilayer-Perceptron (MLP) neural networks. The performances of the models derived by the two approaches are demonstrated using a range of engine tests and by analysing their static and dynamic behaviours.
Application of artificial neural networks to micro gas turbines
2011
In this work, artificial neural networks (ANNs) were applied to describe the performance of a micro gas turbine (MGT). In particular, they were used (i) to complete performance diagrams for unavailable experimental data; (ii) to assess the influence of ambient parameters on performance; and (iii) to analyze and predict emissions of pollutants in the exhausts. The experimental data used to feed the ANNs were acquired from a manufacturer’s test bed. Though large, the data set did not cover the whole working range of the turbine; ANNs and an artificial neural fuzzy interference system (ANFIS) were therefore applied to fill information gaps. The results of this investigation were also used for sensitivity analysis of the machine’s behavior in different ambient conditions. ANNs can effectively evaluate both MGT performance and emissions in real installations in any climate, the worst R2 in the validation set being 0.9962.