A gear system transmits power by means of meshing gear teeth and is conceptually simple and effective in power transmission. Thus typical applications include electric utilities, ships, helicopters, and many other industrial applications. Monitoring the condition of large gearboxes in industries has attracted increasing interest in the recent years owing to the need for decreasing the downtime on production machinery and for reducing the extent of secondary damage caused by failures. This paper addresses the development of a condition monitoring procedure for a gear transmission system using artificial neural networks (ANNs) and support vector machines (SVMs). Seven conditions of the gear were investigated: healthy gear and gear with six stages of depthwise wear simulated on the gear tooth. The features extracted from the measured vibration and sound signals were mean, root mean square (rms), variance, skewness, and kurtosis, which are known to be sensitive to different degrees of faults in rotating machine elements. These characteristics were used as an input features to ANN and SVM. The results show that the multilayer feed forward neural network and multiclass support vector machines can be effectively used in the diagnosis of various gear faults.
To investigate the effect of different proximate index on minimum ignition temperature(MIT) of coal dust cloud, 30 types of coal specimens with different characteristics were chosen. A two-furnace automatic coal proximate analyzer was employed to determine the indexes for moisture content, ash content, volatile matter, fixed carbon and MIT of different types of coal specimens. As the calculated results showed that these indexes exhibited high correlation, a principal component analysis (PCA) was adopted to extract principal components for multiple factors affecting MIT of coal dust, and then, the effect of the indexes for each type of coal on MIT of coal dust was analyzed. Based on experimental data, support vector machine (SVM) regression model was constructed to predicate the MIT of coal dust, having a predicating error below 10%. This method can be applied in the predication of the MIT for coal dust, which is beneficial to the assessment of the risk induced by coal dust explosion (CDE).
Accurate network fault diagnosis in smart substations is key to strengthening grid security. To solve fault classification problems and enhance classification accuracy, we propose a hybrid optimization algorithm consisting of three parts: anti-noise processing (ANP), an improved separation interval method (ISIM), and a genetic algorithm-particle swarm optimization (GA-PSO) method. ANP cleans out the outliers and noise in the dataset. ISIM uses a support vector machine (SVM) architecture to optimize SVM kernel parameters. Finally, we propose the GA-PSO algorithm, which combines the advantages of both genetic and particle swarm optimization algorithms to optimize the penalty parameter. The experimental results show that our proposed hybrid optimization algorithm enhances the classification accuracy of smart substation network faults and shows stronger performance compared with existing methods.
This paper analyses the effectiveness of determining gas concentrations by using a prototype WO3 resistive gas sensor together with fluctuation enhanced sensing. We have earlier demonstrated that this method can determine the composition of a gas mixture by using only a single sensor. In the present study, we apply Least-Squares Support-Vector-Machine-based (LS-SVM-based) nonlinear regression to determine the gas concentration of each constituent in a mixture. We confirmed that the accuracy of the estimated gas concentration could be significantly improved by applying temperature change and ultraviolet irradiation of the WO3 layer. Fluctuation-enhanced sensing allowed us to predict the concentration of both component gases.
The purpose of the work was to predict the selected product parameters of the dry separation process using a pneumatic sorter. From the perspective of application of coal for energy purposes, determination of process parameters of the output as: ash content, moisture content, sulfur content, calorific value is essential. Prediction was carried out using chosen machine learning algorithms that proved to be effective in forecasting output of various technological processes in which the relationships between process parameters are non-linear. The source of data used in the work were experiments of dry separation of coal samples. Multiple linear regression was used as the baseline predictive technique. The results showed that in the case of predicting moisture and sulfur content this technique was sufficient. The more complex machine learning algorithms like support vector machine (SVM) and multilayer perceptron neural network (MPL) were used and analyzed in the case of ash content and calorific value. In addition, k-means clustering technique was applied. The role of cluster analysis was to obtain additional information about coal samples used as feed material. The combination of techniques such as multilayer perceptron neural network (MPL) or support vector machine (SVM) with k-means allowed for the development of a hybrid algorithm. This approach has significantly increased the effectiveness of the predictive models and proved to be a useful tool in the modeling of the coal enrichment process.