Figure 1 - uploaded by Ivan Lorencin
Content may be subject to copyright.
Schematic diagram of the analyzed CCPP.
Source publication
In this paper a genetic algorithm (GA) approach to design of multi-layer perceptron (MLP) for combined cycle power plant power output estimation is presented. Dataset used in this research is a part of publicly available UCI Machine Learning Repository and it consists of 9568 data points (power plant operating regimes) that is divided on training d...
Context in source publication
Context 1
... combined cycle power plant has two identical gas turbines with the same operating parameters and consequentially the same produced power. Steam turbine is composed of high-pressure and low-pressure cylinders mounted on the same shaft, with a note that the low-pressure cylinder is a dual flow symmetrical cylinder, as shown in Figure 1. Cumulative power produced from the analyzed combined cycle power plant in each operating regime can be calculated by using two different approaches: the first is conventional approach, while the second is approach by applying Machine Learning (ML) algorithms in order to estimate electrical power output. ...
Similar publications
![Complexity weights of components [19, 23]](profile/Huynh-Thai-Hoc/publication/353366035/figure/tbl1/AS:1098625987952647@1638944255589/Complexity-weights-of-components-19-23_Q320.jpg)
![General systems characteristics [19, 23]](profile/Huynh-Thai-Hoc/publication/353366035/figure/tbl2/AS:1098625996341248@1638944257157/General-systems-characteristics-19-23_Q320.jpg)
This study presents a modified approach to adjust a software development effort estimation. The AdamOptimizer-based regression model is adopted to adjust and enhance the accuracy of effort estimation. This approach is derived into three phases. The first step deals with the logarithmized formula of effort estimation computed by Function Point Analy...
In this work, a comparative study was carried out between two classification methods: The Multi layer Perceptron Artificial Neural Network (MLP ANN) and the method of classification of the Nearest Neighbors, used in the classification of the diagnosis of breast cancer. The data used in this work were taken from the UCI Machine Learning Repository a...
Citations
... A continuous conditional random feld model contributes to this direction by improving the power plant's performance based on mean squared error (MSE) signifcantly, compared with regression trees and neural network techniques [10]. Implementation of a genetic multilayer perceptron algorithm forecasts higher accurate power outcomes, compared with linear regression and pace regression methodologies [11]. ...
Artificial neural networks (ANNs) have gained prominence among contemporary computing techniques due to their capacity to handle complicated stochastic datasets and nonlinear modelling in combined gas and combined cycle power (COGAS) plants. Researchers, academicians, and stakeholders have been unable to predict, ensure effective operation, and prevent power outages in COGAS due to the nonlinearity. The first implementation of the simultaneous adoption of three types of ANNs using Levenberg–Marquardt (LM), Bayesian regularisation (BR), and scaled conjugate gradient (SCG) configurations for training and assessing a combined cycle power plant output is presented. The dataset used in this research is a 9568-unit full combined cycle power plant basis load dataset, accessible through the public UCI Machine Learning Repository. It incorporates ambient temperature, exhaust vacuum, ambient pressure, and relative humidity as input parameters to predict the electric output power. The most accurate and dependable electric power predictions could be identified for 70% of the total data, of which 6698 were trained, 15% were tested, and 15% were validated (2870). By using the three training techniques, namely, LM, BR, and SCG, the parameterized networks are studied, increasing the number of hidden layers from 20 to 500. The lowest root-mean-square error value for a multilayer perceptron (MLP) architecture is 3.631%, which is lower than the values of 4.17%, 4.35%, and 4.63% for comparable MLP structures (20 to 500), documented in the literature. The LM and BR algorithms outperform SCG. These adopted algorithms could be a cutting-edge application in the power plant industry and other real-world applications for reliable solutions, to satisfy emerging societal needs with environmental benefits.
... Steam turbines are nowadays constituent components of various power plants: conventional steam power plants [1], nuclear power plants [2], cogeneration and combined cycle power plants [3,4], marine propulsion power plants [5,6] and many others [7,8]. The dominant function of steam turbines is the drive of electric generators and electricity production [9]. ...
Exergy analysis results for the Whole observed steam Turbine and each of her cylinders at three loads are presented in this paper. Observation of all cylinders shows that LPC (Low Pressure Cylinder) is the dominant mechanical power producer at the highest observed load, while at partial loads the dominant mechanical power producer is IPC (Intermediate Pressure Cylinder). At Load 100% Whole Turbine produces mechanical power equal to 341.11 MW. IPC is the cylinder with the lowest exergy destruction and the highest exergy efficiency (higher than 95%) at all observed loads. The exergy efficiency of the Whole Turbine (WT) continuously increases during the increase in turbine load (WT exergy efficiency is the lowest at Load 50% and equal to 91.36%, while at the Load 100% WT exergy efficiency is the highest and equal to 92.93%). Analyzed turbine is projected to operate dominantly on the Load 100% because at that load the exergy efficiencies of all cylinders and Whole Turbine are higher than 91%.
... In the electricity production worldwide, steam turbines are inevitable components [1,2]. Steam turbines nowadays can be found in many systems and processes, such as conventional [3], supercritical [4], nuclear [5], cogeneration [6], combined [7], marine [8,9] and other power plants. In all mentioned power plants, steam turbines can be composed of only one or more cylinders. ...
In this paper is performed an isentropic analysis of the entire Intermediate Pressure Cylinder (IPC) and all of his four Segments. Obtained results show that the first Segment (Seg. 1) is the dominant mechanical power producer of all Segments and it produces 16816.70 kW of mechanical power in the real (polytropic) expansion process. Analyzed IPC produces more than half mechanical power of the entire turbine in which he operates (in real expansion process IPC produces mechanical power equal to 58499.48 kW). Isentropic loss and isentropic efficiency of IPC Segments are reverse proportional-Seg. 3 which has the highest isentropic loss simultaneously has the lowest isentropic efficiency (equal to 82.44%), while Seg. 4 which has the lowest isentropic loss has the highest isentropic efficiency (equal to 87.26%). Entire IPC has an isentropic efficiency equal to 87.78%. Any improvements and modifications which can potentially be performed in the observed IPC should firstly be based on the turbine stages mounted inside Seg. 3.
... H. Lu et al. analysed energy quality management for a microenergy network and applied GA to optimise the energy distribution in a tourist area [43]. Some authors [44,45] showed that sorting GA could reduce water pumps' electricity requirements and pollution emissions. A double-injection diesel engine is optimised by using a hybrid model of ANN and GA [46]. ...
The fish processing sector is experiencing increased pressure to reduce its energy consumption and carbon footprint as a response to (a) an increasingly stringent energy regulatory landscape, (b) rising fuel prices, and (c) the incentives to improve social and environmental performance. In this paper, a standalone forecasting computational platform is developed to optimise energy usage and reduce energy costs. This short-term forecasting model is achieved using an artificial neural network (ANN) to predict power and temperature at thirty-minute intervals in two cold rooms of a fish processing plant. The proposed ANN function is optimised by genetic algorithms (GA) with simulated annealing algorithms (SA) to model the relationships between future temperature and power and the system variables affecting them. To assess the accuracy of the proposed method, extensive experiments were conducted using real-world data sets. The results of the experiments indicate that the proposed ANN model performs with higher accuracy than (a) the long short-term memory (LSTM) as an artificial recurrent neural network (RNN) architecture, (b) peephole-LSTM, and (c) the gated recurrent unit (GRU). This paper finds that using GA & SA algorithms; ANN parameters can be optimised. The RMSE obtained by the ANN compared with the second-ranked method GRU was consequently 16% and 4% less for the predicted temperature and power. The results in one particular site demonstrate energy cost savings in the range of 15%–18% after applying the forecast-optimiser approach. The proposed prediction model is used in a fish processing plant for energy management and is scalable to other sites.
... Gas turbine power plants are nowadays widely used in various practical applications. Its dominant usage is evident in mechanical power production, where they operate as independent mechanical power producers [1,2] or as a part of various complex systems for mechanical power and/or heat production [3][4][5]. The high combustion gas temperature at the gas turbine power plant outlet allows heat utilization and efficiency increases for the entire system in which they operate [6,7]. ...
This paper presents thermodynamic and improvement potential analyses of a helium closed-cycle gas turbine power plant (Oberhausen II) and dominant plant components at four loads. DESIGN LOAD represents optimal operating conditions that cannot be obtained in exploitation but can be used as a guideline for further improvements. In real plant exploitation, the highest plant efficiency is obtained at NOMINAL LOAD (31.27%). Considering all observed components, the regenerator (helium-helium heat exchanger) is the most sensitive to the ambient temperature change. An exact comparison shows that the efficiency decrease of an open-cycle gas turbine power plant during load decrease is approximately two and a half or more times higher in comparison to a closed-cycle gas turbine power plant. Plant improvement potential related to all turbomachines leads to the conclusion that further improvement of the most efficient turbomachine (High Pressure Turbine-HPT) will increase whole plant efficiency more than improvement of any other turbomachine. An increase in the HPT isentropic efficiency of 1% will result in an average increase in whole plant efficiency of more than 0.35% at all loads during plant exploitation. In the final part of this research, it is investigated whether the additional heater involvement in the plant operation results in a satisfactory increase in power plant efficiency. It is concluded that in real exploitation conditions (by assuming a reasonable helium pressure drop of 5% in the additional heater), an additional heating process cannot be an improvement possibility for the Oberhausen II power plant.
... Whole system exergy destruction is identical in both calculated ways, which confirm the accuracy of used method. Further research related to the observed waste heat recovery supercritical CO 2 system will be based on the application of various artificial intelligence methods and algorithms [65][66][67][68] with an aim to observe improvement possibilities. The main goal will be to perform optimization by decreasing exergy destruction and increasing exergy efficiency of each component and whole system. ...
In this research is performed an exergy analysis of supercritical CO2 system which uses various waste heat flows from marine diesel engine to produce additional mechanical power. The performed exergy analysis contains whole system as well as each system component individually. The observed system produces useful mechanical power equal to 2299.47 kW which is transferred to the main propulsion propeller shaft. Additionally produced mechanical power by using waste heat only will reduce marine diesel engine fuel consumption and exhaust gas emissions. Main cooler has the highest exergy destruction of all system components and simultaneously the lowest exergy efficiency in the observed system, equal to 32.10% only. One of the possibilities how main cooler exergy efficiency can be increased is by decreasing water mass flow rate through the main cooler and simultaneously by increasing water temperature at the main cooler outlet. Observed system has five heat exchangers which are involved in the CO2 heating process, and it is interesting that the last CO2 heater (exhaust gas waste heat exchanger) increases the CO2 temperature more than all previous four heat exchangers. Whole analyzed waste heat recovery supercritical CO2 system has exergy destruction equal to 2161.68 kW and exergy efficiency of 51.54%. In comparison to a similar CO2 system which uses waste heat from marine gas turbine, system analyzed in this paper has approximately 12% lower exergy efficiency due to much lower waste heat temperature levels (from marine diesel engine) in comparison to temperature levels which occur at the marine gas turbine exhaust.
... Also, steam turbines are essential components of any nuclear steam power plant, regardless of its origin (stationary or marine type) [17,18]. In many complex power plants, such as cogeneration or combined power plants, steam turbines are also inevitable elements [19,20]. ...
This paper presents an analysis and comparison of four steam turbines and their cylinders from four different power plants (marine, conventional, ultra-supercritical and nuclear power plants). The main goal was to find which steam turbine and their cylinders show the best performances, the highest efficiencies, the lowest specific steam consumption and which turbine is the lowest influenced by the ambient temperature change. The highest efficiencies, both isentropic and exergy, are observed in the steam turbine and their cylinders from the ultra-supercritical power plant (whole turbine from ultra-supercritical power plant has an isentropic efficiency equal to 88.36% and exergy efficiency equal to 91.05%). Also, this turbine has the lowest specific steam consumption (7.32 kg/kWh) and exergy parameters of this turbine are the lowest influenced by the ambient temperature change. The worst performance (the lowest efficiencies, high specific steam consumption and the highest sensitivity to the ambient temperature change) show the cylinders and whole turbine from marine propulsion power plant. The same analysis and comparison are also performed for several other steam turbines from four mentioned power plants, so the presented relations and dominant conclusions have general validity. It can be concluded that steam turbines in ultra-supercritical power plants show the best performances in comparison to steam turbines from any other power plant.
... Further research related to the observed steam turbine and its cylinders will be based on various Artificial Intelligence (AI) methods and processes [25][26][27] with an aim of more detail analysis and possible optimization. ...
In this paper is presented an exergy analysis of steam turbine (along with analysis of each cylinder and cylinder part) from ultra-
supercritical power plant. Observation of all the cylinders shows that IPC (Intermediate Pressure Cylinder) is the dominant mechanical
power producer (it produces mechanical power equal to 394.44 MW), it has the lowest exergy loss and simultaneously the highest exergy
efficiency (equal to 95.84%). HPC (High Pressure Cylinder) has a very high exergy efficiency equal to 92.37% what confirms that ultra-
supercritical steam process is very beneficial for the HPC (and whole steam turbine) operation. LPC (Low Pressure Cylinder) is a
dissymmetrical dual flow cylinder, so both of its parts (left and right part) did not produce the same mechanical power, did not have the
same exergy loss, but their exergy efficiency is very similar and in a range of entire LPC exergy efficiency (around 82.5%). Whole observed
steam turbine produces mechanical power equal to 928.03 MW, has exergy loss equal to 93.45 MW and has exergy efficiency equal to
90.85%. The exergy efficiency of the whole analyzed steam turbine is much higher in comparison to other steam turbines from various
conventional power plants.
... Another issue is the poor data distribution due to randomized selection-such as having a large number of unacceptable images in one of the subsets. As the selection between the training and testing sets is performed randomly, the selected training data may conform extremely well to the trained model, but the general model performance may in reality be extremely poor [31]. The K-fold method aims to address this by using the entire available dataset both for training and testing data [32], as visualized in Figure 3. ...
A common issue with X-ray examinations (XE) is the erroneous quality classification of the XE, implying that the process needs to be repeated, thus delaying the diagnostic assessment of the XE and increasing the amount of radiation the patient receives. The authors propose a system for automatic quality classification of XE based on convolutional neural networks (CNN) that would simplify this process and significantly decrease erroneous quality classification. The data used for CNN training consist of 4000 knee images obtained via radiography procedure (KXE) in total, with 2000 KXE labeled as acceptable and 2000 as unacceptable. Additionally, half of the KXE belonging to each label are right knees and left knees. Due to the sensitivity to image orientation of some CNNs, three approaches are discussed: (1) Left-right-knee (LRK) classifies XE based just on their label, without taking into consideration their orientation; (2) Orientation discriminator (OD) for the left knee (LK) and right knee (RK) analyses images based on their orientation and inserts them into two separate models regarding orientation; (3) Orientation discriminator combined with knee XRs flipped to the left or right (OD-LFK)/OD-RFK trains the models with all images being horizontally flipped to the same orientation and uses the aforementioned OD to determine whether the image needs to be flipped or not. All the approaches are tested with five CNNs (AlexNet, ResNet50, ResNet101, ResNet152, and Xception), with grid search and k-fold cross-validation. The best results are achieved using the OD-RFK hybrid approach with the Xception network architecture as the classifier and ResNet152 as the OD, with an average AUC of 0.97 (±0.01).
... It will surely be interesting to investigate the improvement possibilities for each steam turbine and its cylinders. Also, each turbine can be performed various optimizations using conventional [73] or artifi cial intelligence methods and techniques, which show its potential not only in the marine sector [74,75], but also in many other applications and processes [76,77]. ...
This paper presents thermodynamic (energy and exergy) analysis and comparison of two diff erent marine propulsion steam turbines based on their operating parameters from exploitation. The fi rst turbine did not possess steam reheating and had only two cylinders (high-pressure and low-pressure cylinders), while the second turbine possesses steam reheating and has one additional cylinder (intermediate-pressure cylinder). In the literature at the moment, there cannot be found a direct and exact comparison of these two marine steam turbines and their cylinders based on real exploitation conditions. Along with energy and exergy analyses, the research it is investigated the sensitivity of exergy parameters to the ambient temperature change for both turbines and each cylinder. It is also presented the infl uence of the steam reheating process on the energy and exergy effi ciency of the entire power plant. For both observed turbines and their cylinders it is valid that relative losses and effi ciencies (both energy and exergy) are reverse proportional. The operation of an intermediate pressure cylinder from a steam turbine with reheating is the closest to optimal. Due to the diff erent origins of losses considered in energy and exergy analyses, each analysis detects diff erent turbine cylinders as the most problematic ones. The steam reheating process decreases losses and increases effi ciencies (both energy of each turbine cylinder and the whole turbine. The whole turbine with reheating has an energy effi ciency equal to 81.46% and an exergy effi ciency equal to 86.48%, while the whole turbine without reheating has energy and exergy effi ciencies equal to 76.47% and 80.94%, respectively. Exergy parameters of a steam turbine without reheating as well as its cylinders are much more infl uenced by the ambient temperature change in comparison to the steam turbine with reheating and its cylinders. The steam reheating process will increase the efficiency of the whole power plant in real exploitation conditions between 10% and 12%.
