查看更多>>摘要：This paper presents the results of research on identifying moisture inside the walls of buildings with the use of electrical impedance tomography (EIT). The original, complex pixel-oriented ensemble method (POE) was used to solve the inverse, ill-posed problem transforming the input measurements into the tomographic output image pixels. The task of POE is to guarantee reconstructions of a quality that exceeds homogeneous algorithmic methods, no matter what other approaches are used. The presented research used four known, homogeneous machine learning methods: elastic net, linear regression with the least-squares learner (LR-LS), linear regression with SVM learner (LR-SVM) and artificial neural networks (ANN), which were trained to generate output images. All algorithms create pixel-by-pixel reconstructions, meaning that a separate predictive model is trained for each pixel. Then, using the POE algorithm, the best of the four reconstruction methods was adjusted to each pixel of the output image, taking into account the given measurement case. Each measurement results in a different assignment of reconstructive methods to pixels. Since POE can optimise the selection of a method for a given pixel taking into account a specific measurement vector, regardless of how many homogeneous methods will be included in the POE algorithm, the results obtained with POE will always exceed any of the homogeneous methods used. There is the fundamental novelty and original contribution of this research to the general state of knowledge.
查看更多>>摘要：Axle-end grounding system (AGS) is an important assembly of the high-speed railway vehicle, which plays a significant role in protecting the gear box and vehicle bogie from the damage of the working current. We proposed a 3D reconstruction of the AGS friction disk with the active vision approach. The concentric conics are chosen as the feature on the calibration target for the efficient use of available degrees of freedom and the shape similarity of the annular abrasion. The target concentric conics are extracted from the active-contours with the enhanced initial operator to improve the iteration and noise immunity of the conic features under the noise condition. Moreover, the key idea is to derive a novel optimization model established in se(3), which preserves the orthogonality constraint of the measurement system and eliminates the redundancy constraint condition. The experiment and analysis verify that the measurement method provides high accuracy and reconstruction performance.
Cervantes, M. J.Dunca, G.Mulu, B.Jonsson, P. P....
11页
查看更多>>摘要：The paper presents a reformulation of the standard pressure-time method for flow rate measurement in closed conduits. According to the IEC 60041 standard, the method is used with turbines flow rate cut-off. Its current formulation requires complete closure of the turbine shut-off device, generally guide vanes for reaction turbines and valves for impulse turbines, which may cause wear and tear on the machine. Any leakage through the closed shut-off device also needs to be measured. In the present work, a new formulation of the pressure-time method is derived, extending the actual use. The newly formulated method allows the determination of the flow rate at any instant of time from a load variation and, thus, the initial, incremental, and final flow rate. The load variation may be positive or negative. Measurements of any leakage flow are eliminated. Numerical and experimental cases are used to assess the validity and applicability of the proposed formulation. The numerical results demonstrate the correctness of the derivation. The experimental results exemplify the applicability of the method. Limitations of the derived methodology are discussed.
查看更多>>摘要：Joint is a crucial factor affecting rock mass stability, but its complex morphology brings challenges to its research. The three-dimensional (3D) scanning technology can obtain the point cloud digital model of the object without contact. The research is to introduce this technology into the analysis of joint characteristics. For this purpose, firstly, cloud plane inclination of joint 3D scanning data is solved by constructing translation matrix M and rotation matrix R. Then, the appropriate sampling interval for 3D scanning technology to analyze the joint morphology is determined by parametric analysis. Next, by integrating 3D scanning, 3D printing, and 3D carving technologies, the two fabrication methods of joint models with the same natural morphology and lithology are proposed, which overcomes the problem of insufficient joint samples in the experiment. Finally, based on the 3D scanning digital model, the recognition method of the potential shear region and the measurement technique of shear failure characteristics are presented. This research provides a new way to analyze the joint morphology and shear mechanism, and is also conducive to the popularization of 3D scanning technology in geotechnical engineering.
查看更多>>摘要：Temperature distribution on the dam surface plays an important role in reliable thermomechanical stress analysis of arch dams. In this paper, a novel reconstruction method based on a convolutional neural network is proposed for the temperature field on the downstream surface of arch dams. The environmental variables, including solar radiation, mountain shadows, ambient temperature, and weather conditions, are considered in the reconstruction method. The temperature field of the Xiaowan dam is predicted using the proposed method and it agrees well with the monitoring data in the spatial and temporal scale. The results show that the environmental variables significantly affect the temperature on the downstream surface, and result in temporal-spatial variability of nonuniform temperature distribution. In addition, the proposed method solves the data sparseness and lowfrequency acquisition problems that may miss the peak temperature values in the current monitoring system. The thermomechanical stress analysis based on the reconstructed method and finite element method is conducted, and its results illustrate that the proposed reconstructed method can give more realistic performance of the dam and predict possible thermal cracking on the dam surface.
查看更多>>摘要：Convolutional neural network with transfer learning are effective methods for rolling bearing unsupervised learning fault diagnosis. In view of the problem that 1D-CNN cannot give full play to the feature extraction, an improved Adaptive Dimension Convert convolutional neural network (ADC-CNN) is proposed, which can adaptively process one-dimensional vibration signals into two-dimension matrices and input them into 2D-CNN for learning, making full use of the ability of CNN to extract two-dimensional data features. In order to further reduce the data distribution distance between source domain and target domain, the training method of transfer learning is improved by Layered Alternately Transfer Learning (LATL), which layering calculate the CORAL and MK-MMD loss function alternately. To verify the reliability of the proposed method, we carry out experimental verification on the rolling bearing datasets of CWRU and PU. Compared with the traditional 1D-CNN model, the diagnostic classifier accuracy of the proposed ADC-CNN-FLATL is improved by 9% per transfer mission on PU dataset on average, which proves the validity of the proposed method.
查看更多>>摘要：The multimode fiber (MMF) can be adopted to the fiber-coupling laser transmitting optics system of the arrayimaging LiDAR for far-range detection due to the relatively high demand for laser energy which can melt the single-mode fiber. However, the modal noise within the multimode fiber (MMF) can result in severe speckle at the output face which seriously affects the staring imaging LiDAR detection performance. Based on the proposed dual-wavelength LiDAR prototype in which the ICCD detects the 532 nm signal and the array Gm-APD detects the 1064 nm signal, the micro vibrator which is beneficial to the system miniaturization is used for suppressing the speckle. For this application, it's necessary to investigate the speckle suppression effect to evaluate whether the MMF could be suitable for the LiDAR system or not. To our best knowledge, the multimode speckle suppression for dual-wavelength LiDAR hasn't been discussed ever, moreover, the Gm-APD speckle suppression research hasn't been carried out, yet. In this paper, firstly, the principle of the speckle suppression by vibrating the MMF was investigated and the speckle suppression detection models of both detectors were first proposed and discussed. Then, the theoretical analysis was carried out, and it showed that the vibration amplitude of 9 mm and the vibration frequency of higher than 215 Hz could help to get a good suppression results. Finally, the vibrator's parameters were chosen based on the theoretical results, and the experiments were carried out to investigate the effects of the vibrator's voltage which affects the vibration frequency and the exposure time on the image contrast of the intensity and the contrast-to-noise ratio (CNR) for ICCD, besides, the Gm-APD' s contrast of triggering probability was also investigated. Experimental results indicated that the vibrator voltage for getting a good speckle suppression was at least 1.5 V, at which the ICCD's CNR could be enhanced to 1.2 times within 40 ms exposure time and the contrast of intensity of ICCD could be lowered to 0.65 times; besides, the contrast of triggering probability of Gm-APD could be lowered to 0.58 times. Moreover, compared with the speckle suppression results within 5 ms, 40 ms exposure time could help to enhance the CNRs of ICCD to about 3.12 times, lower the contrast of intensity of ICCD and the contrast of the triggering probability of Gm-APD to 0.85 times and 0.84 times, respectively. This paper demonstrates the feasibility of using MMF coupling laser transmitting scheme in the dual-wavelength staring imaging LiDAR, and it provides the experimental and theoretical basis for the system miniaturization and flexible application of dual-wavelength LiDAR.
查看更多>>摘要：The measurement of flow velocity in steep mountainous channels is extremely difficult due to the shallow flow environment. One promising method is capturing the path of tracers and establishing the relation between bulk velocity and tracer transport processes. In this study, two types of dye and salt tracers were utilized for bulk velocity measurement in a steep gravel-bed rectangular channel by analyzing color and electrical conductivity signals. Characteristics velocities gained from the electrical conductivity approach are close to the true values and those from the color space approach need to be corrected. The HSV space signals are more effective to indicate those characteristic boundaries of the dye solution. By analyzing the tracer residence time, longitudinal dispersion characteristics are essentially associated with turbulence negatively and flow resistance linearly and affect the velocity estimation. This study provides new insights to estimate bulk velocity and turbulence induced dispersion in gravel-bed streams.
查看更多>>摘要：This study addresses the maximum queue length problem arises when the arriving traffic exceeds the sensors' detection area, especially for the over-saturated condition. The paper presents an adaptive system working Monte Carlo based Signal Timing (MCaST) algorithm based on microscopic scale vehicle arrival. Since the algorithm works as an adaptive system, two new mathematical formulas for the delay and queue length calculations have been proposed instead for fixed time formulas. Based on both queue length and intersection delay, the intersection performance has been calculated and used as a slave method to optimize the cycle length. The suggested MCaST algorithm has been tested using real field data, and the delay is decreased by 30.2% less during peak hour traffic. Besides, the delay results from the algorithm have been compared with Webster's and Highway Capacity Manual 2010 delay formulas and VISSIM software. Results of the numerical experiments show the MCaST algorithm's validity.
查看更多>>摘要：Shunt capacitance is one of the most critical parts of the test circuit in charge mode, and its influence on impact load measurement of polyvinylidene fluoride (PVDF) is worth discussing. In this study, several experiments are conducted by connecting different shunt capacitances. The results show that in addition to the shunt capacitor, the charge generated is partially concentrated on the PVDF capacitor and extra capacitor, and the amount depends on the capacitance ratio. The source of the extra capacitance is the test circuit parasitic capacitance and the PVDF capacitance change during the impact load, and their influence must be minimized due to difficulty to measure or calculate. Therefore, based on the measurement principles, the interference caused by the extra capacitance in the charge and current modes is analyzed, and suitable measurement conditions are compared. Research in this paper is helpful to improve the accuracy of impact load measurement with PVDF.
NSTL
Elsevier