Search

• A data-driven model is proposed based on multilevel features fusion to classify the variation trend of silicon content. • Multilevel features are extracted through variable analysis, statistical information, and stacked denoising autoencoder, respectively. • Silicon content trend labels are quantitative descripted by seven different primitives. • Both experimental simulation and industrial application verify the effectiveness and feasibility of the proposed model. The silicon content variation trend, which can reflect the quality of molten iron, provides significant information that can assist in ensuring the smooth operation of a blast furnace. This paper proposes a novel dynamic data-driven model for the online classification of the variation trend for the silicon content. Typically, a dynamic model for the silicon content variation trend primarily relies on process data feature extraction. First, a multilevel features fusion algorithm based on mutual information is developed to extract a rich and robust feature representation. Subsequently, the fused multilevel feature vectors and their corresponding trend labels are fed into a recurrent neural network model to capture the process dynamics and classify the variation trend. An experimental simulation and industrial application verified the effectiveness and feasibility of the proposed method. The classification results can provide guidance to ensure that the quality of molten iron is maintained within the desired range in the ironmaking process. [ABSTRACT FROM AUTHOR]

Abstract: Environmental sounds are unstructured and similar to noise. However, the recognition of environmental sounds can benefit crime investigations, warning systems for elderly persons, and security systems. A few past research projects were developed for classifying the environmental sounds. In this paper, we proposed an environmental sound classification algorithm using spectrogram pattern matching along with neural network and k-nearest neighbor (k-NN) classifiers. Unlike other techniques, our approach is based on the observation that local features are more important than global features. In addition, our technique can avoid the problem of filtering less informative and irrelevant frequencies in the classification step. Twenty types of sound from BBC and Sound Ideas databases, with each sound sample longer than 10min, were tested with our algorithm. The spectrogram feature was compared with mel frequency delta cepstral coefficient (MFCC), linear prediction coefficient (LPC), and matching pursuit (MP) features. Two relevant factors concerning the accuracy of classification, window size and sampling rate, were also investigated to find the suitable value of each factor. We also investigated all combinations of these features. Using the k-NN classifier, the maximum accuracy of 94.98% occurred when the spectrogram, LPC, and MP features were combined. The experiments showed that the spectrogram feature with a feed-forward neural network can effectively classify 1-s audio clips with an accuracy of 85.66%. Furthermore, a longer duration can significantly increase the accuracy of classifying each sound clip to a certain limit without affecting cost. Using an audio clip duration of 6s, the spectrogram feature with a feed-forward neural network provided the best classification accuracy, at least 90.57%. [Copyright &y& Elsevier]