Prediction of Electrical Output Power of Combined Cycle Power Plant Using Regression ANN Model

Abstract

Recently, regression artificial neural networks are used to model various systems that have high dimensionality with nonlinear relations. The system under study must have enough dataset available to train the neural network. The aim of this work is to apply and experiment various options effects on feed-foreword artificial neural network (ANN) which used to obtain regression model that predicts electrical output power (EP) of combined cycle power plant based on 4 inputs. Dataset is obtained from an open online source. The work shows and explains the stochastic behavior of the regression neural, experiments the effect of number of neurons of the hidden layers. It shows also higher performance for larger training dataset size; at the other hand, it shows different effect of larger number of variables as input. In addition, two different training functions are applied and compared. Lastly, simple statistical study on the error between real values and estimated values using ANN is conducted, which shows the reliability of the model. This paper provides a quick reference to the effects of main parameters of regression neural networks.

Combined cycle gas turbine power output prediction and data mining with optimized data matching algorithm

“…Table 7 reveals that the prediction performance results fall within quite a narrow range for all the datasets and cases evaluated. The TOB Stage 1 performance is almost identical for all five shuffled datasets, as should be expected (RMSE = 3.53; AAPD % = 0.55; SD = 3.53), significantly outperforming the best of the regression algorithms evaluated by Tufekci [9] and ANN analysis of Elfaki and Hassan [13]. All cases, except Case 5B, show improvements on that accuracy by applying the optimized solutions.…”

“…To place this prediction performance in context, the best of fifteen regression-based machine learning methods evaluated for the same dataset by Tufekci [9], i.e., the Bagging algorithm with REPTree, achieved an RMSE of 3.787 MW. Moreover, the ANN method applied to the dataset by Elfaki & Hassan [13] only achieved an RMSE of 4.32 MW. The remaining algorithms evaluated by Tufekci [9] achieved RMSE's of between 3.861 MW (K-lazy -learning) and 8.487 MW (RBF function).…”

“…A number of studies have applied various regression methods and other machine learning techniques to CCGT performance prediction [9] and fault detection [10]. The application of artificial neural networks (ANN) as reliable alternatives for various simulation control modelling and fault detection applications for CCGT is now well established [11][12][13]. Other machine learning methods including Least Squares Support Vector Machines (LSSVM) [14] and adaptive neuro-fuzzy inference system (ANFIS) [15] are also applied for modelling and controlling GT.…”

“…Here, transparent-open-box (TOB) machine learning [24][25][26] applies an optimized-data-matching technique to model CCGT performance based on a dataset published by UCI [27]. That dataset is already extensively analysed to compare the prediction performance of several correlation and regression-based machine learning methods [9] including ANN [13]. The TOB method, which does not use correlations to make its predictions, offers several advantages over neural network and other correlation machine-learning methods.…”

Optimization of a 660 MWe Supercritical Power Plant Performance—A Case of Industry 4.0 in the Data-Driven Operational Management. Part 2. Power Generation

“…Elfaki et al have modeled the electrical output power of a combined cycle power plant by ANN using four input variables. The deviation between the target value and validation dataset value is negligible, signifying the model's effectiveness [25]. Zhu, H., et al have employed wavelet decomposition and ANN to forecast power generation from the photovoltaic power plant.…”