scholarly journals Application of an artificial neural network for determination of the cylinder-piston group wear level of the automobile-tractor engine

2022 ◽  
Vol 14 (2) ◽  
pp. 62-71
Author(s):  
Andrii Molodan ◽  
◽  
Dmytrii Abramov ◽  
Yurii Tarasov ◽  
Mykola Potapov ◽  
...  
Keyword(s):  
Neural Network ◽  
Artificial Neural Network ◽  
Training Data ◽  
Automotive Engine ◽  
Speed Of Information Processing ◽  
Recognition Result ◽  
Cylinder Piston ◽  
Artificial Neural ◽  
Reliable Knowledge ◽  
Hidden Layer

The article proposes reducing the redundancy of the neural network and the need to reduce the number of neurons in the hidden layer for a given level of network learning error. The minimum number of neurons of the hidden layer for the case of 11 monitoring standard sensors, the parameters of the automobile and tractor engine (ATE) and five classes of typical defects of the ATE nodes can be reduced to 5-7 with a high quality of recognition of the state of the ATE engine. The goal is to provide an expanded reliable knowledge base, the speed of information processing, the accuracy of the resulting technical diagnosis and the ability to quickly determine the technical state of an automotive engine in the mode real time. The basis of the proposed method is to ensure obtaining an extended reliable knowledge base, the speed of information processing, the accuracy of the obtained technical diagnosis and the ability to quickly determine the technical state of an ATE engine in real time. A feature of the proposed method is the use of voltages obtained in an artificial neural network from sensors that are standard in an ATE engine as input signals, and additionally indicate the output signal of the fuel cut-off device, provided for one step, containing a winding of a normally closed electromagnetic valve, which redirects fuel to the drain line. The use of the algorithm for identifying the values of the indicators of operating modes and malfunctions of the cylinder-piston group is the result of the analysis of an artificial neural network, which receive the results of the diagnostic parameters of the automotive engine. Having studied the artificial neural network 1 with different volumes of training data, we obtained the dependence of the change in the reliability of the result on the size of the training data and the reliability of the recognition result is 91.2%, the optimal amount of training data is 1200. Having examined the artificial neural network 2 with different volumes of training data, we obtained the dependence of the change in the reliability the result from the size of the training data and the reliability of the recognition result is 86.5%, the optimal amount of training data is from 10 to 15. The results obtained show the fundamental possibility of creating predictive models of units and assemblies of the tested automotive engines. The model can be created using the apparatus of artificial neural networks and using a fairly large database of tests performed.

Failure pressure prediction of pipeline with single corrosion defect using artificial neural network

2021 ◽  
pp. 166-171
Author(s):  
К. Т. Чин ◽  
Т. Арумугам ◽  
С. Каруппанан ◽  
М. Овинис
Keyword(s):  
Neural Network ◽  
Artificial Neural Network ◽  
Training Data ◽  
Coefficient Of Determination ◽  
Ann Model ◽  
Failure Pressure ◽  
Corrosion Defect ◽  
Artificial Neural ◽  
Artificial Neural Network Ann ◽  
Hidden Layer

Описываются разработка и применение искусственной нейронной сети (ИНС) для прогнозирования предельного давления трубопровода с точечным коррозионным дефектом, подверженного воздействию только внутреннего давления. Модель ИНС разработана на основе данных, полученных по результатам множественных полномасштабных испытаний на разрыв труб API 5L (класс от X42 до X100). Качество работы модели ИНС проверено в сравнении с данными для обучения, получен коэффициент детерминации R = 0,99. Модель дополнительно протестирована с учетом данных о предельном давлении корродированных труб API 5L X52 и X80. Установлено, что разработанная модель ИНС позволяет прогнозировать предельное давление с приемлемой погрешностью. С использованием данной модели проведена оценка влияния длины и глубины коррозионных дефектов на предельное давление. Выявлено, что глубина коррозии является более значимым фактором разрушения корродированного трубопровода. This paper describes the development and application of artificial neural network (ANN) to predict the failure pressure of single corrosion affected pipes subjected to internal pressure only. The development of the ANN model is based on the results of sets of full-scale burst test data of pipe grades ranging from API 5L X42 to X100. The ANN model was developed using MATLAB’s Neural Network Toolbox with 1 hidden layer and 30 neurons. Before further deployment, the developed ANN model was compared against the training data and it produced a coefficient of determination ( R ) of 0.99. The developed ANN model was further tested against a set of failure pressure data of API 5L X52 and X80 grade corroded pipes. Results revealed that the developed ANN model is able to predict the failure pressure with good margins of error. Furthermore, the developed ANN model was used to determine the failure trends when corrosion defect length and depth were varied. Results from this failure trend analysis revealed that corrosion defect depth is the most significant parameter when it comes to corroded pipeline failure.

scholarly journals A neural network for online spike classification that improves decoding accuracy

2020 ◽  
Vol 123 (4) ◽  
pp. 1472-1485
Author(s):  
Deepa Issar ◽  
Ryan C. Williamson ◽  
Sanjeev B. Khanna ◽  
Matthew A. Smith
Keyword(s):  
Neural Network ◽  
Artificial Neural Network ◽  
Real Time ◽  
Target Location ◽  
Training Data ◽  
Electrode Arrays ◽  
Neural Signals ◽  
Network Output ◽  
Artificial Neural ◽  
Hidden Layer

Separating neural signals from noise can improve brain-computer interface performance and stability. However, most algorithms for separating neural action potentials from noise are not suitable for use in real time and have shown mixed effects on decoding performance. With the goal of removing noise that impedes online decoding, we sought to automate the intuition of human spike-sorters to operate in real time with an easily tunable parameter governing the stringency with which spike waveforms are classified. We trained an artificial neural network with one hidden layer on neural waveforms that were hand-labeled as either spikes or noise. The network output was a likelihood metric for each waveform it classified, and we tuned the network’s stringency by varying the minimum likelihood value for a waveform to be considered a spike. Using the network’s labels to exclude noise waveforms, we decoded remembered target location during a memory-guided saccade task from electrode arrays implanted in prefrontal cortex of rhesus macaque monkeys. The network classified waveforms in real time, and its classifications were qualitatively similar to those of a human spike-sorter. Compared with decoding with threshold crossings, in most sessions we improved decoding performance by removing waveforms with low spike likelihood values. Furthermore, decoding with our network’s classifications became more beneficial as time since array implantation increased. Our classifier serves as a feasible preprocessing step, with little risk of harm, that could be applied to both off-line neural data analyses and online decoding. NEW & NOTEWORTHY Although there are many spike-sorting methods that isolate well-defined single units, these methods typically involve human intervention and have inconsistent effects on decoding. We used human classified neural waveforms as training data to create an artificial neural network that could be tuned to separate spikes from noise that impaired decoding. We found that this network operated in real time and was suitable for both off-line data processing and online decoding.

Implementation Issues in Artificial Neural Network Based Thermal Analysis

2000 ◽  
Author(s):  
K. T. Yang
Keyword(s):  
Neural Network ◽  
Artificial Neural Network ◽  
Heat Exchangers ◽  
Training Data ◽  
Data Sets ◽  
Two Phase ◽  
Numerical Experimentation ◽  
Artificial Neural ◽  
Artificial Neural Network Ann ◽  
Hidden Layer

Abstract It is now known the generally it can be demonstrated that artificial neural network (ANN), particularly the fully-connected feedforward configuration with backward propagation error-correction routine, can be a rather effective and accurate tool to correlate performance data of thermal devices such as heat exchangers (Sen and Yang, 2000; Kalogirou, 1999). Good examples are the recent demonstrations for the compact fin-tube heat exchangers (Diaz et al., 1999a; Yang et al., 2000; Pacheco-Vega et al., 1999) including those with complex geometries and also two-phase evaporators (Pacheco-Vega et al., 2000) as well as the dynamic modeling of such heat exchangers and their adaptive control (Diaz et al., 1999b; Diaz et al., 2000). Unfortunately, despite such successes, there are still implementation issues of the ANN analysis which lead to uncertainties in its applications and the achieved results. The present paper discusses such issues and the current practices in dealing with them. Those that will be discussed include the number of hidden layers, the number of nodes in each hidden layer, the range within which the input-output data are normalized, the initial assignment of weights and biases, the selection of training data sets, and the training rate. As will be shown, the specific choices are by no means trivial, and yet are rather important in achieving good ANN results in any given application. Since there are no general sound theoretical basis for such choices at the present time, past experience and numerical experimentation are often the best guides. However, many of these choices and issues relating to them involve optimization. As a result. Some of the existing optimization algorithms may prove to be useful and highly desirable in this regard. The current on-going research to provide some rational basis in these issues will also be discussed. Finally, it will also be mentioned that successfully implemented ANNs have many additional uses in practice. Examples include parameter sensitivity analysis, training, design of new experiments, and clustering of data sets.

scholarly journals Failure pressure prediction of pipeline with single corrosion defect using artificial neural network

2020 ◽  
Vol 4 (1) ◽  
pp. 10-17 ◽  
Author(s):  
Kiu Toh Chin ◽  
◽  
Thibankumar Arumugam ◽  
Saravanan Karuppanan ◽  
Mark Ovinis ◽  
...  
Keyword(s):  
Neural Network ◽  
Artificial Neural Network ◽  
Training Data ◽  
Coefficient Of Determination ◽  
Ann Model ◽  
Failure Pressure ◽  
Corrosion Defect ◽  
Artificial Neural ◽  
Artificial Neural Network Ann ◽  
Hidden Layer

This paper describes the development and application of artificial neural network (ANN) to predict the failure pressure of single corrosion affected pipes subjected to internal pressure only. The development of the ANN model is based on the results of 71 sets of full-scale burst test data of pipe grades ranging from API 5L X42 to X100. The ANN model was developed using MATLAB’s Neural Network Toolbox with 1 hidden layer and 30 neurons. Before further deployment, the developed ANN model was compared against the training data and it produced a coefficient of determination of 0.99. The developed ANN model was further tested against a set of failure pressure data of API 5L X52 and X80 grade corroded pipes. Results revealed that the developed ANN model is able to predict the failure pressure with good margins of error (within 15%). Furthermore, the developed ANN model was used to determine the failure trends when corrosion defect length and depth were varied. Results from this failure trend analysis revealed that corrosion defect depth is the most significant parameter when it comes to corroded pipeline failure.

scholarly journals PEMODELAN JARINGAN SYARAF TIRUAN DENGAN CASCADE FORWARD BACKPROPAGATION PADA KURS RUPIAH TERHADAP DOLAR AMERIKA SERIKAT

2018 ◽  
Vol 7 (1) ◽  
pp. 64-72
Author(s):  
Ekky Rosita Singgih Wigati ◽  
Budi Warsito ◽  
Rita Rahmawati
Keyword(s):  
Neural Network ◽  
Artificial Neural Network ◽  
Processing System ◽  
Training Data ◽  
Neural Network Modeling ◽  
Feed Forward Neural Network ◽  
Output Layer ◽  
Input Layer ◽  
Artificial Neural ◽  
Hidden Layer

Neural Network Modeling (NN) is an information-processing system that has characteristics in common with human brain. Cascade Forward Neural Network (CFNN) is an artificial neural network that its architecture similar to Feed Forward Neural Network (FFNN), but there is also a direct connection from input layer and output layer. In this study, we apply CFNN in time series field. The data used isexchange rate of rupiah against US dollar period of January 1st, 2015 until December 31st, 2017. The best model was built from 1 unit input layer with input Zt-1, 4 neurons in the hidden layer, and 1 unit output layer. The activation function used are the binary sigmoid in the hidden layer and linear in the output layer. The model produces MAPE of training data equal to 0.2995% and MAPE of testing data equal to 0.1504%. After obtaining the best model, the data is foreseen for January 2018 and produce MAPE equal to0.9801%. Keywords: artificial neural network, cascade forward, exchange rate, MAPE 

scholarly journals A neural network for online spike classification that improves decoding accuracy

2019 ◽  
Author(s):  
Deepa Issar ◽  
Ryan C. Williamson ◽  
Sanjeev B. Khanna ◽  
Matthew A. Smith
Keyword(s):  
Neural Network ◽  
Artificial Neural Network ◽  
Real Time ◽  
Target Location ◽  
Training Data ◽  
Electrode Arrays ◽  
Neural Signals ◽  
Network Output ◽  
Artificial Neural ◽  
Hidden Layer

AbstractSeparating neural signals from noise can improve brain-computer interface performance and stability. However, most algorithms for separating neural action potentials from noise are not suitable for use in real time and have shown mixed effects on decoding performance. With the goal of removing noise that impedes online decoding, we sought to automate the intuition of human spike-sorters to operate in real time with an easily tunable parameter governing the stringency with which spike waveforms are classified. We trained an artificial neural network with one hidden layer on neural waveforms that were hand-labeled as either spikes or noise. The network output was a likelihood metric for each waveform it classified, and we tuned the network’s stringency by varying the minimum likelihood value for a waveform to be considered a spike. Using the network’s labels to exclude noise waveforms, we decoded remembered target location during a memory-guided saccade task from electrode arrays implanted in prefrontal cortex of rhesus macaque monkeys. The network classified waveforms in real time, and its classifications were qualitatively similar to those of a human spike-sorter. Compared to decoding with threshold crossings, in most sessions we improved decoding performance by removing waveforms with low spike likelihood values. Furthermore, decoding with our network’s classifications became more beneficial as time since array implantation increased. Our classifier serves as a feasible preprocessing step, with little risk of harm, that could be applied to both offline neural data analyses and online decoding.New & NoteworthyWhile there are many spike-sorting methods that isolate well-defined single units, these methods typically involve human intervention and have inconsistent effects on decoding. We used human classified neural waveforms as training data to create an artificial neural network that could be tuned to separate spikes from noise that impaired decoding. We found that this network operated in real time and was suitable for both offline data processing and online decoding.

scholarly journals The effect of Kurtosis on the accuracy of artificial neural network predictive model

2018 ◽  
Vol 204 ◽  
pp. 02018
Author(s):  
Aisyah Larasati ◽  
Anik Dwiastutik ◽  
Darin Ramadhanti ◽  
Aal Mahardika
Keyword(s):  
Neural Network ◽  
Artificial Neural Network ◽  
Predictive Model ◽  
Predictive Models ◽  
Output Layer ◽  
Input Layer ◽  
Artificial Neural ◽  
Misclassification Rates ◽  
Hidden Layer ◽  
Program Data

This study aims to explore the effect of kurtosis level of the data in the output layer on the accuracy of artificial neural network predictive models. The artificial neural network predictive models are comprised of one node in the output layer and six nodes in the input layer. The number of hidden layer is automatically built by the program. Data are generated using simulation approach. The results show that the kurtosis level of the node in the output layer is significantly affect the accuracy of the artificial neural network predictive model. Platycurtic and leptocurtic data has significantly higher misclassification rates than mesocurtic data. However, the misclassification rates between platycurtic and leptocurtic is not significantly different. Thus, data distribution with kurtosis nearly to zero results in a better ANN predictive model.

scholarly journals Identification Of Number Using Artificial Neural Network Backpropagation

2018 ◽  
Vol 215 ◽  
pp. 01011
Author(s):  
Sitti Amalia
Keyword(s):  
Neural Network ◽  
Artificial Neural Network ◽  
Training Data ◽  
Simulation Test ◽  
Backpropagation Algorithm ◽  
Network Parameter ◽  
Artificial Intelligent ◽  
Design And Implementation ◽  
Network Training ◽  
Artificial Neural

This research proposed to design and implementation system of voice pattern recognition in the form of numbers with offline pronunciation. Artificial intelligent with backpropagation algorithm used on the simulation test. The test has been done to 100 voice files which got from 10 person voices for 10 different numbers. The words are consisting of number 0 to 9. The trial has been done with artificial neural network parameters such as tolerance value and the sum of a neuron. The best result is shown at tolerance value varied and a sum of the neuron is fixed. The percentage of this network training with optimal architecture and network parameter for each training data and new data are 82,2% and 53,3%. Therefore if tolerance value is fixed and a sum of neuron varied gave 82,2% for training data and 54,4% for new data

scholarly journals Modeling the concentration of carbonyl of ethylene propylene diene monomer during the thermal aging using artificial neural network

2021 ◽  
Vol 03 (01) ◽  
pp. 45-52
Author(s):  
Hadjira Maouz ◽  
◽  
Asma Adda ◽  
Salah Hanini ◽  
◽  
...  
Keyword(s):  
Neural Network ◽  
Artificial Neural Network ◽  
Thermal Aging ◽  
Training Data ◽  
Ethylene Propylene Diene Monomer ◽  
Ann Model ◽  
Ethylene Propylene ◽  
Artificial Neural ◽  
Input Variables ◽  
Artificial Neural Network Ann

The concentration of carbonyl is one of the most important properties contributing to the detection of the thermal aging of polymer ethylene propylene diene monomer (EPDM). In this publication, an artificial neural network (ANN) model was developed to predict concentration of carbenyl during the thermal aging of EPDM using a database consisting of seven input variables. The best fitting training data was obtained with the architecture of (7 inputs neurons, 10 hidden neurons and 1 output neuron). A Levenberg Marquardt learning (LM) algorithm, hyperbolic tangent transfer function were used at the hidden and output layer respectively. The optimal ANN was obtained with a high correlation coefficient R= 0.995 and a very low root mean square error RMSE = 0.0148 mol/l during the generalization phase. The comparison between the experimental and calculated results show that the ANN model is able of predicted the concentration of carbonyl during the thermal aging of ethylene propylene diene monomer

Sign in / Sign up

Export Citation Format

Share Document