Classification of Surface Roughness and Distance Measurement Using Artificial Neural Network

In this study we use machine vision to assess surface roughness of machined parts produced by the shaping and milling processes. Machine vision allows for the assessment of surface roughness without touching or scratching the surface, and provides the flexibility for inspecting parts without fixing them in a precise position. The quantitative measures of surface roughness are extracted in the spatial frequency domain using a two-dimensional Fourier transform. Two artificial neural networks, which take roughness features as the input, are developed to determine the surface roughness. The first network is for test parts placed in a fixed orientation, which minimises the deviation of roughness measures. The second network is for test parts placed in random orientations, which gives maximum flexibility for inspection tasks. Experimental results have shown that the proposed roughness features and neural networks are efficient and effective for automated classification of surface roughness.

To promote underwater construction, the tools such as underwater robotic manipulator system, underwater vehicles and so on are necessary. Furthermore, these system need an accurate 3D position parameters. If no position information for automatic control system that system is difficult to be operated. Hence, it is important to take position information. Although 3D position camera can offer 3D position parameters, it has still some deviation. In addition, the manipulator system need accurate distance information for fast and accurate operation. Hence, an accurate distance measurement system is necessary for a remotely teleoperated robotic manipulator system for underwater constructions. This paper focus on the distance measurement study to offer a accurate target distance information. The information can combine with 3D position camera to obtain more accurate 3D position parameters to offer a useful position information for underwater constructions, with a remotely teleoperated robotic manipulator system. In this paper, three distance measurement system, includes commercial ultrasonic distance measurement system, fabricated ultrasonic system and a laser distance measurement system is study. The commercial distance measurement system have a distance measurement system rang 8~35 cm, plusmn0.1 cm. However, a short distance measurement is difficult to be measured with the measurement system. On the other side, the fabricated measurement system can measure a short distance. Information from 0~5 cm, accuracy plusmn0.3 cm. The fabricated distance measurement system fabricated with PZT thin film. In this study, a horn is added to improve the accuracy of distance measurement. In addition, the laser distance measurement system, measurement system is used to measure the distance information also in this study. The combination of these distance measurement system to form a multiple range distance measurement can offer a long measurement range for accurate distance measurement underwater. A detailed experiment is to be studied in the future. Furthermore, the distance measurement system has been offered a distance measurement in a remotely teleoperated robotic manipulator system for underwater constructions

Purpose – This paper aims to study the comparison between a response surface methodology (RSM) and artificial neural network (ANN) in the modelling and prediction of surface roughness during endmilling of glass-fibre-reinforced polymer composites. Design/methodology/approach – Aiming to achieve this goal, several milling experiments were performed with polycrystalline diamond inserts at different machining parameters, namely, feed rate, cutting speed, depth of cut and fibre orientation angle. Mathematical model is created using central composite face-centred second-order in RSM and the adequacy of the model was verified using analysis of variance. ANN model is created using the back propagation algorithm. Findings – With regard to the machining test, it was observed that feed rate is the dominant parameter that affects the surface roughness, followed by the fibre orientation. The comparison results show that models provide accurate prediction of surface roughness in which ANN performs better than RSM. Originality/value – The data predicted from ANN are very nearer to experimental results compared to RSM; therefore, this ANN model can be used to determine the surface roughness for various fibre-reinforced polymer composites and also for various machining parameters.

Artificial neural network (ANN) models have been used in the past to model surface roughness in manufacturing processes. Specifically, different parameters influence surface roughness in fused filament fabrication (FFF) processes. In addition, the characteristics of the networks have a direct impact on the performance of the models. In this work, a study about the use of ANN to model surface roughness in FFF processes is presented. The main objective of the paper is discovering how key ANN parameters (specifically, the number of neurons, the training algorithm, and the percentage of training and validation datasets) affect the accuracy of surface roughness predictions. To address this question, 125 3D printing experiments were conducted changing orientation angle, layer height and printing temperature, and measuring average roughness Ra as response. A multilayer perceptron neural network model with backpropagation algorithm was used. The study evaluates the effect of three ANN parameters: (1) number of neurons in the hidden layer (4, 5, 6 or 7), (2) training algorithm (Levenberg–Marquardt, Resilient Backpropagation or Scaled Conjugate Gradient), and (3) data splitting ratios (70%–15%–15% vs. 55%–15%–30%). Mean Absolute Error (MAE) was used as the performance metric. The Resilient Backpropagation algorithm, 7 neurons, and using 55% of training data yielded the best predictive performance, minimizing the MAE. Additionally, the impact of the dataset size on prediction accuracy was analysed. It was observed that the performance of the ANN gets worse as the number of datasets is reduced, emphasizing the importance of having sufficient data. This study will help to select appropriate values for the printing parameters in FFF processes, as well as to define the characteristics of the ANN to be used to model surface roughness.

In metal cutting, tool failure and surface roughness are the important aspects that affect product quality and production cost, and these are affected mainly by vibration of workpiece. Current techniques do not have a proper method to measure vibration of a rotating workpiece so as to use it as a parameter to replace a cutting tool at an appropriate time. The purpose of the present work is therefore to use of laser Doppler vibrometer (LDV) to measure the vibration of workpiece without interfering the machining. Subsequent to obtaining the workpiece vibration data, artificial neural network (ANN) method was adopted to predict surface roughness and root mean square (RMS) velocity of workpiece vibration. According to Taguchi design of experiments, 18 experiments were prepared with two levels of nose radius and three levels of cutting speed and feed rate. Experiments were conducted on CNC lathe to obtain data of surface roughness and RMS of workpiece vibration velocity in boring of AISI 1040. A multilayer feedforward ANN model was developed and trained with the experimental data using back propagation algorithm. Further, the ANN was used to predict surface roughness and RMS velocity of workpiece vibration. The predicted values were compared with the collected experimental data and percentage error was computed. Less percentage of error was found between the experimental and predicted values.

Artificial Neural Networks are a computational paradigm formed based on the neural structure of intelligent organisms to gain better knowledge. Artificial neural networks are often used for various computing purposes. One of them is for prediction (forecasting) data. The type of artificial neural network that is often used for prediction is the artificial neural network backpropagation because the backpropagation algorithm is able to learn from previous data and recognize the data pattern. So from this pattern backpropagation able to analyze and predict what will happen in the future. In this study, the data to be predicted is Human Development Index data from 2011 to 2015. Data sourced from the Central Bureau of Statistics of North Sumatra. This research uses 5 architectural models: 3-8-1, 3-18-1, 3-28-1, 3-16-1 and 3-48-1. From the 5 models of this architecture, the best accuracy is obtained from the architectural model 3-48-1 with 100% accuracy rate, with the epoch of 5480 iterations and MSE 0.0006386600 with error level 0.001 to 0.05. Thus, backpropagation algorithm using 3-48-1 model is good enough when used for data prediction.

A nonlinear 5 layer artificial neural autoencoder network for image data compression is constructed and trained using the back propagation algorithm and medical CT images. The influence of linear and nonlinear pre/postprocessing operations is studied as well as an alternative compression scheme. Important implementational issues of neural networks are addressed as well as autoencoder issues. One of the results of this work is a compression/decompression tool that provides maximum flexibility and can be used independently from the training environment.

Back Propagation (BP) algorithm is one of the oldest learning techniques used by artificial neural networks (ANN). It has successfully been implemented in various practical problems. However, the algorithm still faces some drawbacks such as getting easily stuck at local minima and needs longer time to converge on an acceptable solution. Recently, the introduction of Second Order Methods has shown a significant improvement on the learning in BP but it still has some drawbacks such as slow convergence and complexity. To overcome these limitations, this research proposed a modified approach for BP by introducing the Conjugate Gradient and Quasi-Newton which were Second Order methods together with ‘gain’ parameter. The performances of the proposed approach is evaluated in terms of lowest number of epochs, lowest CPU time and highest accuracy on five benchmark classification datasets such as Glass, Horse, 7Bit Parity, Indian Liver Patient and Lung Cancer. The results show that the proposed Second Order methods with ‘gain’ performed better than the BP algorithm.

This paper proposes an Artificial Neural Networks (ANN) based technique for transient stability prediction. The ANN makes use of the advent of Phasor Measurements Units (PMU) for real-time prediction. Rate of change of bus voltages and angles is used to train a two layers ANN. Potential of the proposed approach is tested using the Egyptian Power System (EPS) as a study system.

This paper presents a new approach to evaluate the extent and locate the area of focal ischemic injury by using bispectrum estimation of electroencephalograms (EEGs) and artificial neural network (ANN). The graded ischemic injuries in 24 Sprague-Dawley (SD) rats were induced for different periods of 8, 18, 30 min by infusing physiological saline along the different blood streams, based on the model for rat focal ischemic cerebral injury described in this paper. Four channels of EEG were collected in each rat at the scheduled time of ischemia. The maximum magnitude and the weighted center of EEG bispectrum (WCOB) were extracted from the EEG bispectrum and a four layer ANN was employed for prediction. Training and testing the ANN used the 'leave one out' method. The level and location of ischemic injury was verified and classified by observing the ischemic area in the heat shock protein (HSP70) test. The proposed system was able to correctly detect the ischemic extent and simply locate the ischemic area in average accuracy of 91.67% and 100% of the cases, respectively. The results show that the method can be used for the detection of focal cerebral ischemia in clinical prognosis.

Evaluation of neural network based simulation models for the prediction of site specific soil properties of the large agricultural areas will provide important technological benefits for the efficient use of land resources and Agricultural Decision Making System. In this study, an artificial intelligence model was investigated for simulation of site specific iron (Fe) and zinc (Zn) levels in the soils of apple orchard by using a feedforward multilayered Artificial Neural Network (ANN). The measurements of Fe and Zn were made for the topsoil (0-25 cm) and subsoil samples (25-50 cm) collected from forty five different coordinates within intervals of 20 x 10 m based on a grid sampling system in the east and north directions. The measured coordinate values were reflected to the input layer of the developed two input one output feedforward multilayer of the ANN model. Respectively, the measured Fe and Zn values were reflected to its output layer, and then the training stage was started. In the training, back propagation algorithms were used to get the most suitable NN structure for the prediction. As a result of varied trainings, the ANN having two hidden layers and ten neurons produced the best estimation values for site specific Fe and Zn estimations. The most suitable prediction values (R2 = 0.98 and R2 = 0.97, P<0.01) were obtained from the ANN with 10.10.1 structure for site specific Fe and Zn levels in the topsoil. When the simulated values obtained from the ANN model were compared with the measured values, it was observed that successful results were achieved based on varied training values. Hence, the ANN based model should be calibrated for varied training conditions depending on different soil conditions to get more reliable results. Key words: Articifial neural network, neural structures, iron, zinc, apple orchard.

Support vector machine (SVM) model is employed and tested for the soil surface roughness classification. SVM is calibrated (trained) and tested with the experimentally obtained data. The experimentally data is obtained by using X-band (9.5 GHz) scatterometer for two soil surface roughness 3.78 cm and 1.83 cm at constant soil surface moisture equal to 22.80%. The measurement of the scattering coefficient was carried out over a range of incidence angle from 20° to 70° at the step of 5° for both the HH and vv polarization. The performance of the SVM model is evaluated from the outcome classification result on trained data set and test data set. Radial Gaussian kernel function results 100% correct Classification and identification of soil surface roughness both in training and validation phase. SVM is a proficient technique for soil surface roughness classification by such experimentation and have numerous of advantages over artificial neural network (ANN) based approaches and other theoretical approaches as its less complexity and less time consumption ability.

The aim of this study is to develop an artificial neural network (ANN) model for spark assisted diamond face grinding (SADFG) of cobalt bonded tungsten carbide (WC-Co) composite to predict the material removal rate (MRR) and average surface roughness (Ra). The experiments were conducted on a self-developed face grinding setup, which is attached with EDM machine. A bronze metal bonded diamond wheel is used for experimentations. All the experiments were performed according to the central rotatable design. The current, pulse on-time, duty factor and wheel speed were taken as input process parameters and responses are measured in terms of MRR and Ra. Central rotatable design is used for experimentation. The obtained experimental data set was used to train the ANN model. The ANN architecture with back propagation algorithm has been used for modelling of process parameters of SADFG process. It has been found that the developed ANN model is capable to predict the MRR and Ra with absolute average percentage error of 10.40% and 6.81%, respectively. It has been also found that wheel speed at 1,300 RPM is suitable for achieving of the better surface finish while duty factor at 0.70 has been found more appropriate for higher MRR.

This study proposes a surface profile and roughness measurement system for a fibre-optic interconnect based on optical interferometry. On the principle of Fizeau interferometer, an interference fringe is formed on the fibre end-face of the fibre-optic interconnect, and the fringe pattern is analysed using the Fast Fourier transform method to reconstruct the surface profile. However, as the obtained surface profile contains some amount of tilt, a rule for estimating this tilt value is developed in this paper. The actual fibre end-face surface profile is obtained by subtracting the estimated tilt amount from the surface profile, as calculated by the Fast Fourier transform method, and the corresponding surface roughness can be determined. The proposed system is characterized by non-contact measurement, and the sample is not coated with a reflector during measurement. According to the experimental results, the difference between the roughness measurement result of an Atomic Force Microscope (AFM) and the measurement result of this system is less than 3 nm.

Stability analysis of a three-term backpropagation algorithm