Epilepsy as a common disease of the nervous system, with high incidence, sudden and recurrent characteristics. Therefore, timely prediction of seizures and intervention treatment can significantly reduce the accidental injury of patients and protect the life and health of patients. Epilepsy seizures is the result of temporal and spatial evolution, Existing deep learning methods often ignore its spatial features, in order to make better use of the temporal and spatial characteristics of epileptic EEG signals. We propose a CBAM-3D CNN-LSTM model to predict epilepsy seizures. First, we apply short-time Fourier transform(STFT) to preprocess EEG signals. Secondly, the 3D CNN model was used to extract the features of preictal stage and interictal stage from the preprocessed signals. Thirdly, Bi-LSTM is connected to 3D CNN for classification. Finally CBAM is introduced into the model. Different attention is given to the data channel and space to extract key information, so that the model can accurately extract interictal and pre-ictal features. Our proposed approach achieved an accuracy of 97.95%, a sensitivity of 98.40%, and a false alarm rate of 0.017 h ^-1 on 11 patients from the public CHB-MIT scalp EEG dataset.

Rolling element bearings are an important joint in mechanical equipment and have a high engineering application value. To solve the problem of the difficulty in extracting periodic fault pulses due to complex noise interference and the interference of transmission paths in rolling element bearing fault characteristic signals, a novel hybrid fault diagnosis method based on complementary complete ensemble robust local mean decomposition with adaptive noise (CCERLMDAN) combined with multiple disturbance multi-verse optimizer (MDMVO)-based Maximum correlated Kurtosis deconvolution (MCKD) is proposed in this paper, and applied in different rolling element bearing fault conditions. Firstly, the CCERLMDAN method adaptively decomposes the fault vibration signal into multiple product functions (PF), and then selects the PF with the most fault information through the sensitive index (SI). Finally, the MDMVO method adaptively selects the best parameter combination of the MCKD method and then uses MCKD to perform a deconvolution operation on the selected PF, highlighting the periodic fault pulse excited by the bearing fault. The field-measured vibration signals of rolling element bearing faults are applied to verify the proposed method. The final results show that the method effectively improves the fault diagnosis accuracy of rolling element bearings, and both CCERLMDAN and MDMVO methods achieve a better performance than the original method.

The rigid guide is an important part of the coal mine lifting system, serving a guiding function during the cage operation, making guide deformation detection crucial for ensuring safety in coal mine production. To address the issues of high detection safety costs, long maintenance times, and poor detection accuracy in the existing manual inspection mode, this article proposes installing detection devices such as industrial cameras and laser scanners on the hoisting container’s roof. During the operation of the lifting system, rigid guide contour laser stripe images are captured, and a laser and machine vision fusion reconstruction model is derived to complete the 3-D reconstruction of the guide point cloud model and obtain guide deformation parameter information. Building upon this, an adaptive center extraction and smoothing algorithm (ACES-R) for the reflective surface stripes of rigid guides is proposed. This method employs an adaptive dynamic contour tracking algorithm and incorporates laser stripe morphology features to effectively segment the laser stripe region. Under complex lighting conditions, including strong reflection interference, overlapping noisy laser segments, and uneven stripe widths, it accurately extracts guide contour information and reconstructs a complete guide point cloud model, outperforming traditional center extraction algorithms. In guide deformation measurement experiments, multiple repeated measurements show that the average error does not exceed 0.2 mm, and the standard deviation (SD) of the measurement results remains within 0.1 mm.

To address the issue of error accumulation arising from the continuous rotation of polarizers within the polarization rotation table, this paper proposes a computational methodology for determining the true rotation angle of polarizers in instances where the rotation angle is unknown. From obtaining atmospheric polarization data, we delve into the relationship between the incident and outgoing light as it traverses optical components, assuming knowledge of the initial polarization state. Furthermore, we expound on the transformational relation between Stokes vectors and Mueller matrices. We leverage these foundational theories to attain precise atmospheric polarization information for multiple pixels. To mitigate the cumulative errors arising from the continuous rotation of polarizers, we use a nonlinear least squares optimization approach to determine polarization rotation angles in the images accurately. Additionally, this innovative approach was compared with the Gauss-Newton method. Simulation results clearly show that, under identical conditions, the algorithm described in this paper surpasses the Gauss-Newton method regarding precision, iteration count, and efficiency. Experimental validation unequivocally affirms the efficacy of the proposed algorithm in precisely estimating polarization rotation angles in images, a pivotal component for the subsequent meticulous reconstruction of atmospheric polarization patterns.