MATLAB Neural Network Toolbox Research Articles

FE and ANN model of ECS to simulate the pipelines suffer from internal corrosion

As the study of internal corrosion of pipeline need a large number of experiments as well as long time, so there is a need for new computational technique to expand the spectrum of the results and to save time. The present work represents a new non-destructive evaluation (NDE) technique for detecting the internal corrosion inside pipeline by evaluating the dielectric properties of steel pipe at room temperature by using electrical capacitance sensor (ECS), then predict the effect of pipeline environment temperature (o) on the corrosion rates by designing an efficient artificial neural network (ANN) architecture. ECS consists of number of electrodes mounted on the outer surface of pipeline, the sensor shape, electrode configuration, and the number of electrodes that comprise three key elements of two dimensional capacitance sensors are illustrated. The variation in the dielectric signatures was employed to design electrical capacitance sensor (ECS) with high sensitivity to detect such defects. The rules of 24-electrode sensor parameters such as capacitance, capacitance change, and change rate of capacitance are discussed by ANSYS and MATLAB, which are combined to simulate sensor characteristic. A feed-forward neural network (FFNN) structure are applied, trained and tested to predict the finite element (FE) results of corrosion rates under room temperature, and then used the trained FFNN to predict corrosion rates at different temperature using MATLAB neural network toolbox. The FE results are in excellent agreement with an FFNN results, thus validating the accuracy and reliability of the proposed technique and leads to better understanding of the corrosion mechanism under different pipeline environmental temperature.

Online performance evaluation of a heavy-duty automatic transmission launching process

To improve the repeatability of control effectiveness of automatic transmission launching process, the method of online evaluation of a heavy-duty automatic transmission launching process is studied, which deals with not only the evaluation of launching quality, but also the refining direction of control parameters when launching quality is not well enough. First, automatic transmission launching process and its key control parameters are anatomized. Second, overall evaluation and phase evaluation indices are proposed in light of launching quality and ideal control process respectively. Third, evaluation model based on BPNN (Error Back Propagation Neural Network) is built and trained using Matlab neural network toolbox. The trained evaluation model is then ported to MCU (Micro Controller Unit) to implement online evaluation. Finally, vehicle tests are carried out. The results indicate that the developed online evaluation method of automatic transmission launching process has a high accuracy and successfully substitutes mechanical evaluation for manual evaluation.

Artificial neural network application to predict the sawability performance of large diameter circular saws

To predict the performance of a large diameter circular saw (LDCS) is among the fundamental steps that are required for determining the practicability of stone production. Natural stone processing plants were visited to measure the areal slab production rate (ASPR) of LDCS in different operational conditions. Neural network toolbox in MATLAB is applied in order to develop a model to predict ASPR of LDCS. An artificial neural network is trained with physical and mechanical properties of eleven stones as input parameters and their associated ASPR values as the target. Uniaxial compressive strength (UCS), Brazilian tensile strength (BTS), Cerchar abrasivity index (CAI), porosity, and density are the physical and mechanical properties that are used as input parameters. In view of its speed, robustness, and the fact that it is very well renowned compared to the other learning algorithms, the Levenberg–Marquardt propagation algorithm is used to train the network. It is explained in detail that a neural network with the previously mentioned input parameters and only one hidden-layer can successfully estimate ASPR for LDCS. It is noticed that, while the number of neurons is less than eight in the single hidden-layer, the network generalizes better than when the number of neurons increases. However, beyond that point, not only the number of neurons does not have any positive effect on performance of the network, but it may also cause the network to memorize the results instead of generalizing them. It can be declared that using ANN to predict ASPR of LDCS may lead the engineers toward a more reliable design and planning.

SYNTHESIS OF NEURAL NETWORK MODEL REFERENCE CONTROLLER FOR AIMING AND STABILIZING SYSTEM

The aim of this work is the synthesis of neural network reference model controller. The synthesis is performed in MATLAB for the problem of control of the aiming and stabilization system for the special equipment of moving objects. This paper presents the synthesis of the neural network reference model controller to meet the given performance characteristics of operation for the aiming and stabilization system for the special equipment of moving objects. Simulink tool in MATLAB is used to build the block diagram of double-loop neural network system of aiming and stabilization, where the reference model controller is put in the velocity loop and P-regulator is put in the position loop, with feedforward velocity control. Presented the method of synthesis of the neural network reference model controller that is implemented in the Neural Network Toolbox in MATLAB. System tests with the broad range of parameter values determined the key parameters defining the control quality. Optimal values of the key parameters were found to provide the highest control performance. System simulation and analysis of the obtained results is given .

Permeability Prediction of Tight Sandstone Reservoirs Using Improved BP Neural Network

By analyzing the permeability controlling factors of tight sandstone reservoir in Wuhaozhuang Oil Field, the permeability is considered to be mainly controlled by porosity, clay content, irreducible water saturation and diagenetic coefficient. Because the conventional BP algorithm has its drawbacks such as slow convergence speed and easy falling into the local minimum value, an improved three-layer feed-forward BP neural network model is built by MATLAB neural network toolbox to predict permeability according to the four permeability controlling factors, while studying samples of model are selected based on the representative core analysis data. The simulation based on improved neural network model shows that the improved model has a faster convergence speed and better accuracy. The consistency between model prediction value and lab test value is good and the mean squared error is less. Therefore, the new model can meet the needs of the development geology research of oil field better in the future.

Comparison Study and Result Analysis of Improved Back-Propagation Algorithms in Bangla Speech Recognition

This research is concerned with the study of different improved and faster back-propagation (BP) algorithms of neural networks and the analysis of recognition result in continuous Bangla speech. For speech recognition, a comparison study on neural networks and speech recognition result analysis with different improved and faster BP algorithms (such as, BP with momentum, variable learning rate BP, resilient BP, conjugate gradient BP and Levenberg-Marquardt BP algorithms) have been done. In this research, the MATLAB Neural Network Toolbox 7.12.0 is used to create, train and simulate the feedforward neural network with the BP learning algorithm. The convergence obtained from standard BP algorithm is very slow; that's why, this research proposes different improved and faster BP algorithms to solve the speech recognition problems. The developed system has been justified by several networks trained with different Bangla speech words. To test the performance of the system, 20 samples of 50 Bangla speech words have been used; from which 10 samples of 50 words are used as training pattern and another 10 samples of 50 words are used as testing pattern in the network. The binary features of speech words have been generated using dynamic thresholding algorithm. The recognition system has been achieved recognition rate of 83% using resilient BP algorithm, 90% using conjugate gradient BP algorithm and 90% using Levenberg-Marquardt BP algorithm, respectively, for recognising 50 speech words.

Ship Performance Research Based on BP Neural Network

The problem is solved that it is hard to provide analysis formulas about the maximum equivalent stress, the maximum shear stress and the structural geometric parameters for a ship. The finite element calculation is done with orthogonal experimental design under the most dangerous case. The data obtained are used as the training and test samples to establish BP neural network models of ship’s maximum equivalent stress and maximum shear stress. With the aid of Neural network toolbox in MATLAB, the topological structure of BP neural network mapping relationship between the whole ship performance indexes and design variables is established. The training and testing are completed with the data tested by the shipyard and the correctness of this network is verified. The neural network required for further optimization design is obtained. The neural network is helpful in reducing the ship mass without exceeding the allowable stress.

Application of Forecasting Power Based on Neural Network and Wilcoxon Rank Sum Test

With the development of modern power systems, the accuracy requirement for load forecasting is getting higher and higher. In this paper, the Wilcoxon rank sum test as a new mathematical method to test error, is applied to the neural network based short-term load forecasting model. The actual historical load data and the associated weather conditions factors to be considered, based on MATLAB neural network toolbox, The method to construct three different neural network model, the power load in Yichang area in 2009 was forecasted and the simulation of short-term. The simulation results show：Wilcoxon rank sum test algorithms in neural network model, not only can correctly predict，relative to the average relative error, in the statistical methods and statistical laws - especially in the face of large amounts of data - through a random sample to show a certain advantage in the overall distribution.

Part Family Formation Based on Clustering Algorithm and BP Network

To solve the problem of design and manufacturing on the production of many varieties of small batch, the parts grouping method that based on clustering algorithm, clustering validity index and BP neural network method for new parts is proposed. At first, mathematical model of part clustering is built, and parts grouping is based on the similarity of data and cluster centers which is calculated by Euclidean distance, then the effectiveness of parts group is tested by Function Index and optimal number of clusters group can be found. The algorithm is achieved by Matlab clustering toolbox, so the best part family structure is built. Furthermore, the grouped parts are used to train the BP neural network toolbox in Matlab, then simulate new parts on network to find the match group. At last, a case study was also presented to verify the feasibility of this method．

On-Line Handwriting Recognition Based on Hopfield Neural Network

In this paper, Discrete Hopfield Neural Network (DHNN) is adopted to realize handwritten characters recognition. First, learning samples are preprocessed including binarization, normalization and interpolation. Then pixel features are extracted and used to establish DHNN. The handwritten test samples and noise corrupted samples are finally inputted into the network to verify its recognition performance. Simulation results reveal that DHNN has good fault tolerance and disturbance rejection performance. In addition, the recognition system is realized with MATLAB neural network toolbox and GUI, which verifies the feasibility of the algorithm.

Comparison of artificial neural network and multiple linear regression models to predict optimum bonding strength of heat treated woods

In this study, an artificial neural network (ANN) model was developed for predicting an optimum bonding strength of heat treated woods. The MATLAB Neural Network Toolbox was used for the training and optimization of the ANN model. The ANN model having the best prediction performance was detected by trying various networks. Then, the ANN results were compared with the results of multiple linear regression (MLR) model. It was shown that the ANN model produced more successful results compared to MLR model in all cases. The mean absolute percentage errors (MAPE) were found as 1.49% and 3.06% in the prediction of bonding strength values for training and testing data sets, respectively. Determination coefficient (R2) values for training and testing data sets in the prediction of bonding strength by ANN were 0.997 and 0.986, respectively. The results also indicated that the designed model is a useful, reliable and quite effective tool for optimizing the effects of heat treatment conditions on bonding strength of wood. Thanks to using optimum bonding strength values obtained by the model, the increase of the bonding quality of wood products can be provided and the costs for experimental material and energy can be reduced.

Neural Network Modeling For Main Steam Temperature System

Main Steam Temperature (MST) is non-linear, large inertia, long dead time and load dependant parameters. The paper present MST modeling method using actual plant data by utilizing MATLAB's Neural Network toolbox. The result of the simulation showed the MST model based on actual plant data is possible but the raw data need to be pre-processed for better output. Generator output, total main steam flow, main steam pressure and total spray flow are four main parameters affected the behavior of MST in coal fired power plant boiler.

Objects Centroid Correlation Using MATLAB's Neural Network Toolbox for Visually-Guided Object Manipulation

Since a pick-and-place task plays an important role in an automatic process, it normally requires machine vision to locate an object for grasping. This paper presents a practicable method used to visually guide an object grasping a group of small, 1.1 mm diameter, screws by using an inexpensive webcam with a resolution of 640 x 480. A basic feedforward neural network is utilized to make a fitting function which associates pixel coordinates of the camera to the physical coordinates of the robot while the method of linear least squares is used for comparison in parallel. The result from the feedforward neural network shows that fifty screws can be completely manipulated from a tray after their physical coordinates are loaded into the robot while the result from the method of linear least squares shows failure when picking two of the samples.

Measurement and evaluation of surface roughness based on optic system using image processing and artificial neural network

The traditional devices, used to measure the surface roughness, are very sensitive, and they are obtained by scratching the surface of materials. Therefore, the optic systems are used as alternatives to these devices to avoid the unwanted processes that damage the surface. In this study, face milling process was applied to American Iron and Steel Institute (AISI) 1040 carbon steel and aluminium alloy 5083 materials using the different tools, cutting speeds and depth of cuts. After these processes, surface roughness values were obtained by the surface roughness tester, and the machined surface images were taken using a polarise microscope. The obtained images were converted into binary images, and the images were used as input data to train network using the MATLAB neural network toolbox. For the training networks, log-sigmoid function was selected as transfer function, scaled conjugate gradient (SCG) algorithm was used as training algorithm, and performance of the trained networks was achieved as an average of 99.926 % for aluminium alloy (AA) 5083 aluminium and as an average of 99.932 % for AISI 1040 steel. At the end of the study, a prediction programme for optical surface roughness values using MATLAB m-file and GUI programming was developed. Then, the prediction programme and neural network performance were tested by the trial experiments. After the trial experiments, surface roughness values obtained with stylus technique for the carbon steel and aluminium alloy materials were compared with the developed programme values. When the developed programme values were compared with the experimental results, the results were confirmed each other at a rate of 99.999 %.

Investigation on trophic state index by artificial neural networks (case study: Dez Dam of Iran)

Dam construction and surface runoff control is one of the most common approaches for water-needs supply of human societies. However, the increasing development of social activities and hence the subsequent increase in environmental pollutants leads to deterioration of water quality in dam reservoirs and eutrophication process could be intensified. So, the water quality of reservoirs is now one of the key factors in operation and water quality management of reservoirs. Hence, maintaining the quality of the stored water and identification and examination of changes along time has been a constant concern of humans that involves the water authorities. Traditionally, empirical trophic state indices of dam reservoirs often defined based on changes in concentration of effective factors (nutrients) and its consequences (increase in chlorophyll a), have been used as an efficient tool in the definition of dam reservoirs quality. In recent years, modeling techniques such as artificial neural networks have enhanced the prediction capability and the accuracy of these studies. In this study, artificial neural networks have been applied to analyze eutrophication process in the Dez Dam reservoir in Iran. In this paper, feed forward neural network with one input layer, one hidden layer and one output layer was applied using MATLAB neural network toolbox for trophic state index (TSI) analysis in the Dez Dam reservoir. The input data of this network are effective parameters in the eutrophication: nitrogen cycle parameters and phosphorous cycle parameters and parameters that will be changed by eutrophication: Chl a, SD, DO and the output data is TSI. Based on the results from estimation of modified Carlson trophic state index, Dez Dam reservoir is considered to be eutrophic in the early July to mid-November and would be mesotrophic with decrease in temperature. Therefore, a decrease in water quality of the dam reservoir during the warm seasons is expectable. The results indicated that artificial neural network (ANN) is a suitable tool for quality modeling of reservoir of dam and increment and decrement of nutrients in trend of eutrophication. Therefore, ANN is a suitable tool for quality modeling of reservoir of dam.

Neural‐network‐based space‐vector pulse‐width modulation for capacitor voltage balancing of three‐phase three‐level improved power quality converter

In this study, an artificial neural-network (ANN)-based space-vector pulse-width modulation (SVPWM) for capacitor voltage balancing of a three-phase three-level neutral-point clamped converter with improved power quality is presented. The neural-network-based controller offers the advantage of very fast implementation of the SVPWM algorithm. This makes it possible to use an application specific integrated circuit chip in place of a digital signal processor. The proposed scheme employs single layer feed-forward neural-networks at different stages along with a control algorithm using modified reference vector for capacitor voltage balancing of an improved power quality three-phase neutral-point clamped converter. In other words, the neural-network receives three-phase voltages and currents as input and generates symmetrical pulse-width modulation waves for three phases of the converter. A simulated digital signal processor (DSP)-based modulator generates the data which are used to train the network by a back-propagation algorithm in the MATLAB Neural Network Toolbox. The simulation of converter with ANN-based space-vector modulator shows excellent performance when compared with that of conventional DSP-based modulator.

Based on RBF Neural Network Modeling of Well Test System

The Neural Network Toolbox in MATLAB is a powerful instrument of analyzing and designing a neural network system. RBF Neural Network has small computational burden and fast learning rate and is not liable to be trapped by local minimal points etc. So it is an effective means to identify and model a system. In this paper, the Neural Network Toolbox in MATLAB and RBF Neural Network are combined to solve the problem of modeling the pressure in oilfield test well systems and the result is excellent.

ANN Modeling of Nickel Base Super Alloys for Time Dependent Deformation

Alloys 617 and 276 are nickel-based super alloys with excellent mechanical properties, oxidation, creep- resistance, and phase stability at high temperatures. These alloys are used in complex and stochastic applications. Thus, it is difficult to predict their output characteristics mathematically. Therefore, the non-conventional methods for modeling become more effective. These two alloys have been subjected to time-dependent deformation at high temperatures under sustained loading of different values. The creep results have been used to develop the new models. Artificial neural network (ANN) was applied to predict the creep rate and the anelastic elongation for the two alloys. The neural network contains twenty hidden layer with feed forward back propagation hierarchical. The neural network has been designed with MATLAB Neural Network Toolbox. The results show a high correlation between the predicted and the observed results which indicates the validity of the models. 

Artificial Neural Networks Modeling of Release of Olanzapine From Glycerol Monooleate Matrices

The important aim of this research was to develop an appropriate model to predict relationships between three casual factors on the responses based on an artificial neural network (ANN). As model formulation, 28 types of gels were prepared. The weight ratio of GMO/water (w/w) and PEG 300/GMO (w/w), percentage of Olanzapine (OZ) were selected as input data. Entrapment efficacy, maximum percentage of release, particle size and viscosity were estimated as gel characterization. A set of gel characterization and input data were employed as tutorial data for ANN methodology by using neural network toolbox in Matlab. Different topologies have been performed in order to determine the single network with good performance and accuracy. Four training algorithms (Levenberg–Marquardt, Bayesian- Regularization, BFGS Quasi-Newton, and Gradient Descent) were applied to train ANNs containing different numbers of hidden layers with various nods. The ability to predict the responses of all the algorithms were in the order of: BR > LM >BFGS> GD.

Flank wears Simulation by using back propagation neural network when cutting hardened H-13 steel in CNC End Milling

High speed milling has many advantages such as higher removal rate and high productivity. However, higher cutting speed increase the flank wear rate and thus reducing the cutting tool life. Therefore estimating and predicting the flank wear length in early stages reduces the risk of unaccepted tooling cost. This research presents a neural network model for predicting and simulating the flank wear in the CNC end milling process. A set of sparse experimental data for finish end milling on AISI H13 at hardness of 48 HRC have been conducted to measure the flank wear length. Then the measured data have been used to train the developed neural network model. Artificial neural network (ANN) was applied to predict the flank wear length. The neural network contains twenty hidden layer with feed forward back propagation hierarchical. The neural network has been designed with MATLAB Neural Network Toolbox. The results show a high correlation between the predicted and the observed flank wear which indicates the validity of the models.