The SnO 2 /ZnO hierarchical heterostructures (HHSs) were synthesized via the microwave-assisted hydrothermal method, and the SnO 2 /ZnO HHSs based sensor exhibited ultra-low detection limit of 2 ppb for detecting NO 2 .
Both non-metallic g-C3N4 and BiVO4 are novel photocatalysts responsive to visible light, but their low charge separation efficiency restricts their inconspicuous photocatalytic activity. In this paper, direct Z-type g-C3N4/BiVO4 photocatalyst was constructed by calcination and hydrothermal for the degradation of methylene blue. The existence of g-C3N4/BiVO4 heterojunction was confirmed by the detailed study of its chemical structure and morphology by various characterization methods, such as X-ray diffraction (XRD), Scanning electron microscope (SEM), and X-ray photoelectron spectroscopy (XPS). The evaluation of photocatalytic performance showed that the MB degradation performance of 1.0-CN/BVO was significantly enhanced, which was 4.528 times and 2.387 times higher than pristine BiVO4 and g-C3N4, respectively, which was mainly due to the enhanced light capture ability and effective electron transfer in the photocatalytic reaction. The 1.0-CN/BVO composite exhibited extremely catalytic stability and recyclability.
30 August 2018 in Journal of Materials Science: Materials in Electronics
For the three-dimensional (3D) point cloud acquisition system, accurately estimating the measurement model parameters is essential, so a calibration algorithm based on the kernel mean p -power error (KMPE) cost function for the 3D point cloud acquisition system is proposed. Firstly, a space sphere with unknown radius is chosen as the calibration target. Secondly, since the KMPE cost function is insensitive to measurement noise and outliers, we establish a KMPE cost function-based nonlinear optimization model for the measurement model parameters by using the geometric constraints of the scanning points on the sphere. Thirdly, combining the successful history-based differential evolution parameter adaptation (SHADE) algorithm with the Levenberg-Marquardt (LM) algorithm to optimize the optimization model, and the optimal measurement model parameters can be obtained. Experimental findings confirm that the 3D point cloud acquisition system can be calibrated precisely by using the proposed algorithm, which can also significantly reduce the impact of the measurement noise and outliers.
Aiming at solving the estimation problem of measurement model parameters of 3D laser scanner, a kernel mean p-power error (KMPE) loss function-based calibration algorithm is proposed in this paper. The KMPE loss function is robust to measurement noise and outliers. Thus, a KMPE loss function-based nonlinear optimization model of measurement model parameters of 3D laser scanner is firstly established according to the distance constraint between each laser scanning point on the calibration sphere and the center of the calibration sphere. Then, to conduct calibration of the model parameters, the optimization model is solved by combining the success-history based parameter adaptation for differential evolution (SHADE) algorithm with the Levenberg-Marquardt (LM) algorithm. Experimental results show that the proposed algorithm can achieve high calibration accuracy of the measurement model parameters of the 3D laser scanner by effectively suppressing the influence of measurement noise and outliers.
1 September 2021 in Journal of Physics: Conference Series
Because the number of students not attending school is expanding at an alarming pace, and because of the COVID-19 epidemic, 102 countries have implemented nationwide closures to conduct local shut-downs and temporarily close schools. This slowed down learning possibilities and intellectual growth even more. Every country’s equity disparities may widen. As a result, we must restructure our educational system so that students can gain correct knowledge and teachers can track how much each student has learned. As a result, Machine Learning and Deep Learning are the most effective solutions for this type of problem. As a result, we’re releasing a method for tackling this challenge using Tabnet, Transformers, LGBM, and a variety of other machine learning approaches for student knowledge tracing.