During advanced water detection using the transient electromagnetic method, the exploration effect for water-rich area is often poor due to the interference of bolts that are distributed in different positions in work...During advanced water detection using the transient electromagnetic method, the exploration effect for water-rich area is often poor due to the interference of bolts that are distributed in different positions in working face. Thus, the study on the interference characteristics of bolts in different states has important directive significance for improving the acquisition quality and data processing method in water detection. Based on the analysis of the distribution laws of magnetic field excited by small multi-turn coincident loop in full space of homogeneity, the test on the interference of bolts has been designed in the mine. Through drilling 18 holes around the overlapping coil in the working face, mass data are collected in order with the posi- tion change and the exposed bolt length. The results of comprehensive data analysis show that the transient electromagnetic field is strongly interfered as the distance between the bolt and the center of the coil is less than 3 m, and the interference varies greatly as the distance varies. On the other hand, the field induced by the bolts can be ignored as the distance exceeds 3 m. The findings can help to improve data acquisition and correction during advanced water detection when using the transient electromagnetic method.展开更多
A 40-60 t/h modularized dry coal beneficiation process with a novel method to control the bed was designed around a gas-solid fluidized bed separator. Furthermore, the hydrodynamics of medium-solids consisting of wide...A 40-60 t/h modularized dry coal beneficiation process with a novel method to control the bed was designed around a gas-solid fluidized bed separator. Furthermore, the hydrodynamics of medium-solids consisting of wide-size-range magnetite powder (0.3-0.06 ram) and 〈1 mm fine coal were numerically studied. The simulation results show that the fluidization performance of the wide-size-range medium-solid bed is good. The separation performance of the modularized system was then investigated in detail using a mixture of 〈0.3 mm magnetite powder (mass fraction of 0.3-0.06 mm particles is 91.38 %) and 〈1 mm fine coal as solid media. The experimental results show that at separation densities of 1.33 g/cm^3 or 1.61 g/cm^3, 50-6 mm coal can be separated effectively with probable error, E, values of 0.05 g/cm^3 and 0.06 g/cm^3, respectively. This technique is beneficial for saving water resources and for the clean utilization of coal.展开更多
An approach to design multi-channel cylinder dryer was proposed. The heat transfer performance and flow characteristic under various structural parameters were analyzed. First, an experiment was designed and set up to...An approach to design multi-channel cylinder dryer was proposed. The heat transfer performance and flow characteristic under various structural parameters were analyzed. First, an experiment was designed and set up to measure the condensing heat transfer coefficient and the pressure drop in order to ,~erify the applicability of the Cavallini's correlation. Then, the relationship among the count of channels, aspect ratio, spacing ratio, width, height and hydraulic diameter of a channel was given. Finally, the correlation of condensing heat transfer and the homogeneous model was introduced in order to observe the heat transfer performance and flow characteristic of the multi-channel cylinder dryer affected by different structures. The study reveals that the structural parameters including count of channels, aspect ratio, spacing ratio of a channel dramatically influence the condensation heat transfer coefficient and frictional resistance of the steam. Based on the selected paper machine, it is suggested that the overall performance of the multi-channel cylinder dryer is best if the count of channels is 150-200, the aspect ratio is 1 : 3 and the spacing ratio is 1 : 1-1 : 3.展开更多
Artificial neural networks have dramatically improved the performance of many machine-learning applications such as image recognition and natural language processing. However, the electronic hardware implementations o...Artificial neural networks have dramatically improved the performance of many machine-learning applications such as image recognition and natural language processing. However, the electronic hardware implementations of the above-mentioned tasks are facing performance ceiling because Moore’s Law is slowing down. In this article, we propose an optical neural network architecture based on optical scattering units to implement deep learning tasks with fast speed, low power consumption and small footprint.The optical scattering units allow light to scatter back and forward within a small region and can be optimized through an inverse design method. The optical scattering units can implement high-precision stochastic matrix multiplication with mean squared error < 10-4 and a mere 4*4 um2 footprint.Furthermore, an optical neural network framework based on optical scattering units is constructed by introducing "Kernel Matrix", which can achieve 97.1% accuracy on the classic image classification dataset MNIST.展开更多
基金Supported by the Key Projects of Anhui Provincial Scientific and Technological Program (11010401015) the Key Program of National Natural Science Foundation of China (51134012)
文摘During advanced water detection using the transient electromagnetic method, the exploration effect for water-rich area is often poor due to the interference of bolts that are distributed in different positions in working face. Thus, the study on the interference characteristics of bolts in different states has important directive significance for improving the acquisition quality and data processing method in water detection. Based on the analysis of the distribution laws of magnetic field excited by small multi-turn coincident loop in full space of homogeneity, the test on the interference of bolts has been designed in the mine. Through drilling 18 holes around the overlapping coil in the working face, mass data are collected in order with the posi- tion change and the exposed bolt length. The results of comprehensive data analysis show that the transient electromagnetic field is strongly interfered as the distance between the bolt and the center of the coil is less than 3 m, and the interference varies greatly as the distance varies. On the other hand, the field induced by the bolts can be ignored as the distance exceeds 3 m. The findings can help to improve data acquisition and correction during advanced water detection when using the transient electromagnetic method.
基金Projects(50921002, 50774084) supported by the National Natural Science Foundation of ChinaProject(2007AA05Z318) supported by the National High-tech Research and Development Program of China+1 种基金Project(BK2010002) supported by the Natural Science Foundation of Jiangsu Province of ChinaProject(20100480473) supported by the China Postdoctoral Science Foundation
文摘A 40-60 t/h modularized dry coal beneficiation process with a novel method to control the bed was designed around a gas-solid fluidized bed separator. Furthermore, the hydrodynamics of medium-solids consisting of wide-size-range magnetite powder (0.3-0.06 ram) and 〈1 mm fine coal were numerically studied. The simulation results show that the fluidization performance of the wide-size-range medium-solid bed is good. The separation performance of the modularized system was then investigated in detail using a mixture of 〈0.3 mm magnetite powder (mass fraction of 0.3-0.06 mm particles is 91.38 %) and 〈1 mm fine coal as solid media. The experimental results show that at separation densities of 1.33 g/cm^3 or 1.61 g/cm^3, 50-6 mm coal can be separated effectively with probable error, E, values of 0.05 g/cm^3 and 0.06 g/cm^3, respectively. This technique is beneficial for saving water resources and for the clean utilization of coal.
基金Acknowledgements This project is supported by the National Natural Science Foundation of China (Grant No. 51375286), Natural Science Basic Research Plan in Shaanxi Province of China (Program No. 2012JZ7002) and the key scientific and technological innovation team fund of Shaanxi Province of China (Program No. 2014KCT- 15).
文摘An approach to design multi-channel cylinder dryer was proposed. The heat transfer performance and flow characteristic under various structural parameters were analyzed. First, an experiment was designed and set up to measure the condensing heat transfer coefficient and the pressure drop in order to ,~erify the applicability of the Cavallini's correlation. Then, the relationship among the count of channels, aspect ratio, spacing ratio, width, height and hydraulic diameter of a channel was given. Finally, the correlation of condensing heat transfer and the homogeneous model was introduced in order to observe the heat transfer performance and flow characteristic of the multi-channel cylinder dryer affected by different structures. The study reveals that the structural parameters including count of channels, aspect ratio, spacing ratio of a channel dramatically influence the condensation heat transfer coefficient and frictional resistance of the steam. Based on the selected paper machine, it is suggested that the overall performance of the multi-channel cylinder dryer is best if the count of channels is 150-200, the aspect ratio is 1 : 3 and the spacing ratio is 1 : 1-1 : 3.
基金This work was supported by the National Key Research and Development Program of China(2017YFA0205700)the National Natural Science Foundation of China(61927820)Yurui Qu was supported by Zhejiang Lab’s International Talent Fund for Young Professionals.
文摘Artificial neural networks have dramatically improved the performance of many machine-learning applications such as image recognition and natural language processing. However, the electronic hardware implementations of the above-mentioned tasks are facing performance ceiling because Moore’s Law is slowing down. In this article, we propose an optical neural network architecture based on optical scattering units to implement deep learning tasks with fast speed, low power consumption and small footprint.The optical scattering units allow light to scatter back and forward within a small region and can be optimized through an inverse design method. The optical scattering units can implement high-precision stochastic matrix multiplication with mean squared error < 10-4 and a mere 4*4 um2 footprint.Furthermore, an optical neural network framework based on optical scattering units is constructed by introducing "Kernel Matrix", which can achieve 97.1% accuracy on the classic image classification dataset MNIST.