Prediction of induced draft fan power consumption in 500MW steam generators using artificial neural network

2017 ◽  
Author(s):  
A. Sriram ◽  
P. R. Venkateswaran ◽  
Sishaj P. Simon
Keyword(s):  
Neural Network ◽  
Artificial Neural Network ◽  
Power Consumption ◽  
Steam Generators ◽  
Artificial Neural ◽  
Induced Draft Fan

Influence of Various Cutting Fluids on Energy Consumption during Turning

2016 ◽  
Vol 686 ◽  
pp. 252-256
Author(s):  
Michal Bachraty ◽  
Marian Tolnay ◽  
Pavel Kovač ◽  
Vladimir Pucovsky
Keyword(s):  
Neural Network ◽  
Artificial Neural Network ◽  
Energy Consumption ◽  
Power Consumption ◽  
Cutting Fluid ◽  
Cutting Fluids ◽  
Parameter Combination ◽  
Artificial Neural ◽  
Behavior Function ◽  
Machining Parameter

In this paper an experiment was carried out to examine the magnitude of differences among cutting fluids and their influence on lathe power consumption during machining. It was discovered that there is no universal cutting fluid. An attempt was made to study the possibility of Artificial Neural Network to model the behavior function for all cutting fluids. This could be used as a foundation for later database building where it would be possible to predict how certain cutting fluid will behave in a specific machining parameter combination.

Modeling Machinability Parameters of Turning Al-SiC (10p) MMC by Artificial Neural Network

2008 ◽  
Author(s):  
N. Muthukrishnan ◽  
Ravi Mohan ◽  
M. S. Thiagarajan ◽  
J. Venugopal
Keyword(s):  
Neural Network ◽  
Artificial Neural Network ◽  
Power Consumption ◽  
Metal Matrix ◽  
Specific Power ◽  
Specific Power Consumption ◽  
Feed Forward ◽  
Ann Models ◽  
Artificial Neural ◽  
Cutting Speeds

The paper presents the results of an experimental investigation on the machinability of fabricated Aluminum metal matrix composite (A356/SiC/10p) during continuous turning of composite rods using medium grade Polycrystalline Diamond (PCD 1500) inserts. Metal Matrix Composites (MMC’s) are very difficult to machine and PCD tools are considered by far, the best choice for the machining of these materials. Experiments were conducted at LMW-CNC-LAL-2 production lathe using PCD 1500 grade insert at various cutting conditions and parameters such as surface roughness and specific powers consumed were measured. The present results reaffirm the suitability of PCD for machining MMCs. Though BUE formation was observed at low cutting speeds, at high cutting speeds very good surface finish and low specific power consumption could be achieved. An Artificial Neural Network (ANN) model has been developed for prediction of machinability parameters of MMC using feed forward back propagation algorithm. The various stages in the development of ANN models VIZ. selection of network type, input and output of the network, arriving at a suitable network configuration, training of the network, validation of the resulting network has been taken up. A 2-9-9-2 feed forward neural network has been successfully trained and validated to act as a model for predicting the machining parameters of Al-SiC (10p) -MMC. The ANN models after successful training are able to predict the surface quality; and specific power consumption for a given set of input values of cutting speed and machining time.

An Artificial Neural Network Model for Monitoring Real-Time Parameters and Detecting Early Warnings in Induced Draft Fan

2018 ◽  
Author(s):  
Di Hu ◽  
Gang Chen ◽  
Tao Yang ◽  
Cheng Zhang ◽  
Ziwen Wang ◽  
...  
Keyword(s):  
Neural Network ◽  
Artificial Neural Network ◽  
Real Time ◽  
Ann Model ◽  
Distribution Law ◽  
Restricted Boltzmann Machines ◽  
Time Parameters ◽  
Artificial Neural ◽  
Induced Draft Fan ◽  
Early Warnings

This paper describes a method to monitor real time parameters and detect early warnings in induced draft fan (ID FAN). An artificial neural network (ANN) model based on cross-relationships among operating parameters was established. In particular, this paper adopted the pre-training of Restricted Boltzmann machines (RBM) and analyzed the training errors of model. A new approach was proposed to monitor parameters by predicted value of model and distribution law of training error, and the reasonable range of each parameter was defined to detect the early warnings in real time. Combining the historical operational data of the No. 1 induced draft fan of No. 3 generating unit in Shajiao C Power Plant in China, this work used MATLAB to verify and analyze the proposed method. The numerical examples shown that the proposed method has better detection performance than the fixed upper and lower limits in the safety instrumented system (SIS). Moreover, this work can expand to other machinery that could be used in manufacturing easily.

Power Consumption Modeling for Indoor Environment Using Artificial Neural Network

2013 ◽  
Vol 315 ◽  
pp. 221-225 ◽  
Author(s):  
Ahmad F.A. Rahman ◽  
Hazlina Selamat ◽  
Fatimah S. Ismail
Keyword(s):  
Neural Network ◽  
Artificial Neural Network ◽  
Power Consumption ◽  
Indoor Environment ◽  
Electrical Power ◽  
Simulated Data ◽  
Ann Model ◽  
Actual System ◽  
Artificial Neural ◽  
Artificial Neural Network Ann

In this paper, a new Artificial Neural Network (ANN) model that relates human comfort and electrical power consumption of a building with temperature, illumination and carbon dioxide (CO2) level inside the building is developed. The model has been developed using samples of simulated data representing the indoor environment variables. Results have shown that neural network with 14 hidden layer neurons produces outputs that is closest to the actual system outputs.

scholarly journals Multilayer Neural Network with Synapse Based on Two Successive Memristors

2018 ◽  
Vol 12 (1) ◽  
pp. 132-147
Author(s):  
Minh-Huan Vo
Keyword(s):  
Neural Network ◽  
Artificial Neural Network ◽  
Power Consumption ◽  
Character Recognition ◽  
Accuracy Rate ◽  
Parity Problem ◽  
Positive Weights ◽  
Artificial Neural ◽  
Memristor Synapse

Introduction: Synapse based on two successive memristors builds the synaptic weights of the artificial neural network for training three-bit parity problem and five-character recognition. Methods: The proposed memristor synapse circuit creates positive weights in the range [0;1], and maps it to range [-1;1] to program both the positive and negative weights. The proposed scheme achieves the same accuracy rate as the conventional bridge synapse schemes which consist of four memristors. Results and Conclusion: However, proposed synapse circuit decreases 50% the number of memristors and 76.88% power consumption compared to the conventional bridge memristor synapse.

Sign in / Sign up

Export Citation Format

Share Document