Road extraction based on deep learning is one of hot spots of semantic segmentation in the past decade.In this work,we proposed a framework based on codec network for automatic road extraction from remote sensing imag...Road extraction based on deep learning is one of hot spots of semantic segmentation in the past decade.In this work,we proposed a framework based on codec network for automatic road extraction from remote sensing images.Firstly,a pre-trained ResNet34 was migrated to U-Net and its encoding structure was replaced to deepen the number of network layers,which reduces the error rate of road segmentation and the loss of details.Secondly,dilated convolution was used to connect the encoder and the decoder of network to expand the receptive field and retain more low-dimensional information of the image.Afterwards,the channel attention mechanism was used to select the information of the feature image obtained by up-sampling of the encoder,the weights of target features were optimized to enhance the features of target region and suppress the features of background and noise regions,and thus the feature extraction effect of the remote sensing image with complex background was optimized.Finally,an adaptive sigmoid loss function was proposed,which optimizes the imbalance between the road and the background,and makes the model reach the optimal solution.Experimental results show that compared with several semantic segmentation networks,the proposed method can greatly reduce the error rate of road segmentation and effectively improve the accuracy of road extraction from remote sensing images.展开更多
采用KCl作为添加剂,根据差示扫描量热仪（DSC）测得的DSC曲线,对非等温动力学微分方程采用Achar-Brindley-Sharp-Wendworth法拟合实验数据,逻辑选择确定KCl-Na OH混合碱熔分解锆英砂的最可几微分机制函数及动力学参数,并对碱熔分解过程...采用KCl作为添加剂,根据差示扫描量热仪（DSC）测得的DSC曲线,对非等温动力学微分方程采用Achar-Brindley-Sharp-Wendworth法拟合实验数据,逻辑选择确定KCl-Na OH混合碱熔分解锆英砂的最可几微分机制函数及动力学参数,并对碱熔分解过程进行动力学分析。研究结果表明：KCl-Na OH混合碱熔分解锆英砂在分解深度为0.01~0.09范围内时,最可几微分机制函数为f（a）=（1-a）2,表观活化能和指前因子分别为199.7 k J·mol-1和1×1010.39s-1。当分解深度为0.29~0.60时,最可几微分机制函数转变为f（a）=3/2[（1-a）-1/3-1]-1,表观活化能和指前因子转变为139.25 k J·mol-1和1×108.52s-1。KCl的加入改变了碱熔分解反应的表观活化能和指前因子,使得碱熔分解反应表观活化能降低,碱熔体系的反应速率增大。KCl-Na OH混合碱熔分解锆英砂反应在609~665℃时,为化学反应控速。随着碱熔分解反应的继续进行,当反应温度为730~811℃时,锆英砂表面不断被产物层包裹,反应机制转变为三维扩散,球形对称,扩散控制过程。展开更多
根据差示扫描量热仪(DSC)测得的DSC曲线,对非等温动力学微分方程采用Achar-Brindley-Sharp-Wendworth法拟合实验数据,逻辑选择确定锆英砂碱熔分解反应的最可几微分机制函数。采用扫描电子显微镜(SEM)对锆英砂及碱熔分解产物的表面形貌...根据差示扫描量热仪(DSC)测得的DSC曲线,对非等温动力学微分方程采用Achar-Brindley-Sharp-Wendworth法拟合实验数据,逻辑选择确定锆英砂碱熔分解反应的最可几微分机制函数。采用扫描电子显微镜(SEM)对锆英砂及碱熔分解产物的表面形貌进行分析。研究结果表明:锆英砂碱熔分解反应在分解深度为0.01~0.10范围内时,最可几微分机制函数为f(a)=(1-a)2,表观活化能和指前因子分别为245.42 k J·mol^(-1)和1×10^(13.2)s-1。当分解深度为0.28~0.66时,最可几微分机制函数转变为f(a)=3/2[(1-a)^(-1/3)-1]^(-1),表观活化能和指前因子转变为163.90 k J·mol^(-1)和1×10~7.1s^(-1)。锆英砂碱熔分解反应最初先在锆英砂的外表面发生反应,为化学反应控速。随着碱熔分解反应的进行,锆英砂表面不断被产物层包裹,越来越厚的产物层对反应物质点扩散的阻碍作用增加,氢氧化钠的质点必须扩散到锆英砂颗粒内部进行反应。此时,反应阻力主要来源于扩散,反应机制转变为三维扩散,球形对称,扩散控制过程。展开更多
基于OpenGL与MFC提出了一种三维图形接口技术,该技术采用UG Open C API编写功能模块,采用OpenGL与MFC开发三维图形接口模块,并利用回调函数机制实现功能模块与三维图形接口模块间的通信。实例证明该技术可扩充UG OPEN C API在界面制作...基于OpenGL与MFC提出了一种三维图形接口技术,该技术采用UG Open C API编写功能模块,采用OpenGL与MFC开发三维图形接口模块,并利用回调函数机制实现功能模块与三维图形接口模块间的通信。实例证明该技术可扩充UG OPEN C API在界面制作和三维图形渲染等方面的功能。展开更多
The random delays in a networked control system (NCS) degrade control performance and can even destabilize the control system.To deal with this problem,the time-stamped predictive functional control (PFC) algorithm is...The random delays in a networked control system (NCS) degrade control performance and can even destabilize the control system.To deal with this problem,the time-stamped predictive functional control (PFC) algorithm is proposed,which generalizes the standard PFC algorithm to networked control systems with random delays.The algorithm uses the time-stamp method to estimate the control delay,predicts the future outputs based on a discrete time delay state space model,and drives the control law that applies to an NCS from the idea of a PFC algorithm.A networked control system was constructed based on TrueTime simulator,with which the time-stamped PFC algorithm was compared with the standard PFC algorithm.The response curves show that the proposed algorithm has better control performance.展开更多
A new general robust fuzzy approach was presented to control the position and the attitude of unmanned flying vehicles(UFVs). Control of these vehicles was challenging due to their nonlinear underactuated behaviors. T...A new general robust fuzzy approach was presented to control the position and the attitude of unmanned flying vehicles(UFVs). Control of these vehicles was challenging due to their nonlinear underactuated behaviors. The proposed control system combined great advantages of generalized indirect adaptive sliding mode control(IASMC) and fuzzy control for the UFVs. An on-line adaptive tuning algorithm based on Lyapunov function and Barbalat lemma was designed, thus the stability of the system can be guaranteed. The chattering phenomenon in the sliding mode control was reduced and the steady error was also alleviated. The numerical results, for an underactuated quadcopter and a high speed underwater vehicle as case studies, indicate that the presented adaptive design of fuzzy sliding mode controller performs robustly in the presence of sensor noise and external disturbances. In addition, online unknown parameter estimation of the UFVs, such as ground effect and planing force especially in the cases with the Gaussian sensor noise with zero mean and standard deviation of 0.5 m and 0.1 rad and external disturbances with amplitude of 0.1 m/s2 and frequency of 0.2 Hz, is one of the advantages of this method. These estimated parameters are then used in the controller to improve the trajectory tracking performance.展开更多
Long waves such as tsunamis can be trapped by islands due to wave refraction,and these trapped waves will cause huge damage even in the sheltered shoreline of the island.That all waves propagating into the topography ...Long waves such as tsunamis can be trapped by islands due to wave refraction,and these trapped waves will cause huge damage even in the sheltered shoreline of the island.That all waves propagating into the topography and finally reaching the coastline are called perfect trapped modes,while any waves escaping from the topography are called leaky modes.Whether these long waves can be trapped is dependent on the depth profile of the island.This paper presents analytic solutions of the ray path for waves propagating into the circular island with power function profiles.Wave height distributions over the island are further investigated based on the principia that crowded rays correspond to large wave height and sparse rays correspond to small wave height.The trapped mechanism for water waves over the island is revealed based on their ray paths.Furthermore,the perfectly trapped criterion is derived,that is,when the slope gradient at the topography toe is greater than twice the ratio of the water depth to the radial distances,all wave rays propagating on the island will finally reach the coastline,and the waves are perfectly trapped.展开更多
On the basis of Preston hypothesis,the motion relationship between tool and workpiece upon the tools motion in planar model is analyzed.The effect on computer controlled optical surfacing (CCOS) caused by controllable...On the basis of Preston hypothesis,the motion relationship between tool and workpiece upon the tools motion in planar model is analyzed.The effect on computer controlled optical surfacing (CCOS) caused by controllable variable is simulated except for the dwelling time,thus,some reference on theory is provided to optimize the former numerical control (NC) model,and fast manufacturing of large departure aspherics is realized.展开更多
A new method called RS-MSVM （Rough Set and Multi-class Support Vector Machine） is proposed for network intrusion detection. This method is based on rough set followed by MSVM for attribute reduction and classificati...A new method called RS-MSVM （Rough Set and Multi-class Support Vector Machine） is proposed for network intrusion detection. This method is based on rough set followed by MSVM for attribute reduction and classification respectively, The number of attributes of the network data used in this paper is reduced from 41 to 30 using rough set theory. The kernel function of HVDM-RBF （Heterogeneous Value Difference Metric Radial Basis Function）, based on the heterogeneous value difference metric of heterogeneous datasets, is constructed for the heterogeneous network data. HVDM-RBF and one-against-one method are applied to build MSVM. DARPA （Defense Advanced Research Projects Agency） intrusion detection evaluating data were used in the experiment. The testing results show that our method outperforms other methods mentioned in this paper on six aspects： detection accuracy, number of support vectors, false positive rate, falsc negative rate, training time and testing time.展开更多
By introducing the distribution of the light energy density in GaN-based light-emitting diode (LED),theLED model based on the incoherent regime and the light extraction efficiency are investigated.The energy density a...By introducing the distribution of the light energy density in GaN-based light-emitting diode (LED),theLED model based on the incoherent regime and the light extraction efficiency are investigated.The energy density asa function of the angle of incidence is calculated to demonstrate the mechanism of the light extraction.The deviationbetween the tendencies of the transmissivity of the output layer and the extraction efficiency is also demonstrated.展开更多
基金supported by National Natural Science Foundation of China(No.61864025)2021 Longyuan Youth Innovation and Entrepreneurship Talent(Team),Young Doctoral Fund of Higher Education Institutions of Gansu Province(No.2021QB-49)+4 种基金Employment and Entrepreneurship Improvement Project of University Students of Gansu Province(No.2021-C-123)Intelligent Tunnel Supervision Robot Research Project(China Railway Scientific Research Institute(Scientific Research)(No.2020-KJ016-Z016-A2)Lanzhou Jiaotong University Youth Foundation(No.2015005)Gansu Higher Education Research Project(No.2016A-018)Gansu Dunhuang Cultural Relics Protection Research Center Open Project(No.GDW2021YB15).
文摘Road extraction based on deep learning is one of hot spots of semantic segmentation in the past decade.In this work,we proposed a framework based on codec network for automatic road extraction from remote sensing images.Firstly,a pre-trained ResNet34 was migrated to U-Net and its encoding structure was replaced to deepen the number of network layers,which reduces the error rate of road segmentation and the loss of details.Secondly,dilated convolution was used to connect the encoder and the decoder of network to expand the receptive field and retain more low-dimensional information of the image.Afterwards,the channel attention mechanism was used to select the information of the feature image obtained by up-sampling of the encoder,the weights of target features were optimized to enhance the features of target region and suppress the features of background and noise regions,and thus the feature extraction effect of the remote sensing image with complex background was optimized.Finally,an adaptive sigmoid loss function was proposed,which optimizes the imbalance between the road and the background,and makes the model reach the optimal solution.Experimental results show that compared with several semantic segmentation networks,the proposed method can greatly reduce the error rate of road segmentation and effectively improve the accuracy of road extraction from remote sensing images.
文摘采用KCl作为添加剂,根据差示扫描量热仪（DSC）测得的DSC曲线,对非等温动力学微分方程采用Achar-Brindley-Sharp-Wendworth法拟合实验数据,逻辑选择确定KCl-Na OH混合碱熔分解锆英砂的最可几微分机制函数及动力学参数,并对碱熔分解过程进行动力学分析。研究结果表明：KCl-Na OH混合碱熔分解锆英砂在分解深度为0.01~0.09范围内时,最可几微分机制函数为f（a）=（1-a）2,表观活化能和指前因子分别为199.7 k J·mol-1和1×1010.39s-1。当分解深度为0.29~0.60时,最可几微分机制函数转变为f（a）=3/2[（1-a）-1/3-1]-1,表观活化能和指前因子转变为139.25 k J·mol-1和1×108.52s-1。KCl的加入改变了碱熔分解反应的表观活化能和指前因子,使得碱熔分解反应表观活化能降低,碱熔体系的反应速率增大。KCl-Na OH混合碱熔分解锆英砂反应在609~665℃时,为化学反应控速。随着碱熔分解反应的继续进行,当反应温度为730~811℃时,锆英砂表面不断被产物层包裹,反应机制转变为三维扩散,球形对称,扩散控制过程。
文摘根据差示扫描量热仪(DSC)测得的DSC曲线,对非等温动力学微分方程采用Achar-Brindley-Sharp-Wendworth法拟合实验数据,逻辑选择确定锆英砂碱熔分解反应的最可几微分机制函数。采用扫描电子显微镜(SEM)对锆英砂及碱熔分解产物的表面形貌进行分析。研究结果表明:锆英砂碱熔分解反应在分解深度为0.01~0.10范围内时,最可几微分机制函数为f(a)=(1-a)2,表观活化能和指前因子分别为245.42 k J·mol^(-1)和1×10^(13.2)s-1。当分解深度为0.28~0.66时,最可几微分机制函数转变为f(a)=3/2[(1-a)^(-1/3)-1]^(-1),表观活化能和指前因子转变为163.90 k J·mol^(-1)和1×10~7.1s^(-1)。锆英砂碱熔分解反应最初先在锆英砂的外表面发生反应,为化学反应控速。随着碱熔分解反应的进行,锆英砂表面不断被产物层包裹,越来越厚的产物层对反应物质点扩散的阻碍作用增加,氢氧化钠的质点必须扩散到锆英砂颗粒内部进行反应。此时,反应阻力主要来源于扩散,反应机制转变为三维扩散,球形对称,扩散控制过程。
文摘The random delays in a networked control system (NCS) degrade control performance and can even destabilize the control system.To deal with this problem,the time-stamped predictive functional control (PFC) algorithm is proposed,which generalizes the standard PFC algorithm to networked control systems with random delays.The algorithm uses the time-stamp method to estimate the control delay,predicts the future outputs based on a discrete time delay state space model,and drives the control law that applies to an NCS from the idea of a PFC algorithm.A networked control system was constructed based on TrueTime simulator,with which the time-stamped PFC algorithm was compared with the standard PFC algorithm.The response curves show that the proposed algorithm has better control performance.
文摘A new general robust fuzzy approach was presented to control the position and the attitude of unmanned flying vehicles(UFVs). Control of these vehicles was challenging due to their nonlinear underactuated behaviors. The proposed control system combined great advantages of generalized indirect adaptive sliding mode control(IASMC) and fuzzy control for the UFVs. An on-line adaptive tuning algorithm based on Lyapunov function and Barbalat lemma was designed, thus the stability of the system can be guaranteed. The chattering phenomenon in the sliding mode control was reduced and the steady error was also alleviated. The numerical results, for an underactuated quadcopter and a high speed underwater vehicle as case studies, indicate that the presented adaptive design of fuzzy sliding mode controller performs robustly in the presence of sensor noise and external disturbances. In addition, online unknown parameter estimation of the UFVs, such as ground effect and planing force especially in the cases with the Gaussian sensor noise with zero mean and standard deviation of 0.5 m and 0.1 rad and external disturbances with amplitude of 0.1 m/s2 and frequency of 0.2 Hz, is one of the advantages of this method. These estimated parameters are then used in the controller to improve the trajectory tracking performance.
基金supported by the National Key Research and Development Program of China(No.2016YFC 1402800)the National Science Fund for Distinguished Young Scholars(No.51425901)+1 种基金the National Natural Science Foundation of China(No.51579090)Innovation Project of Colleges and Universities in Jiangsu Province(No.2015B41814)
文摘Long waves such as tsunamis can be trapped by islands due to wave refraction,and these trapped waves will cause huge damage even in the sheltered shoreline of the island.That all waves propagating into the topography and finally reaching the coastline are called perfect trapped modes,while any waves escaping from the topography are called leaky modes.Whether these long waves can be trapped is dependent on the depth profile of the island.This paper presents analytic solutions of the ray path for waves propagating into the circular island with power function profiles.Wave height distributions over the island are further investigated based on the principia that crowded rays correspond to large wave height and sparse rays correspond to small wave height.The trapped mechanism for water waves over the island is revealed based on their ray paths.Furthermore,the perfectly trapped criterion is derived,that is,when the slope gradient at the topography toe is greater than twice the ratio of the water depth to the radial distances,all wave rays propagating on the island will finally reach the coastline,and the waves are perfectly trapped.
文摘On the basis of Preston hypothesis,the motion relationship between tool and workpiece upon the tools motion in planar model is analyzed.The effect on computer controlled optical surfacing (CCOS) caused by controllable variable is simulated except for the dwelling time,thus,some reference on theory is provided to optimize the former numerical control (NC) model,and fast manufacturing of large departure aspherics is realized.
基金Supported by the 863 High Tech. Project (2001AA140213) and the State Key Basic Research Pro-ject (2001CB309403).
文摘A new method called RS-MSVM （Rough Set and Multi-class Support Vector Machine） is proposed for network intrusion detection. This method is based on rough set followed by MSVM for attribute reduction and classification respectively, The number of attributes of the network data used in this paper is reduced from 41 to 30 using rough set theory. The kernel function of HVDM-RBF （Heterogeneous Value Difference Metric Radial Basis Function）, based on the heterogeneous value difference metric of heterogeneous datasets, is constructed for the heterogeneous network data. HVDM-RBF and one-against-one method are applied to build MSVM. DARPA （Defense Advanced Research Projects Agency） intrusion detection evaluating data were used in the experiment. The testing results show that our method outperforms other methods mentioned in this paper on six aspects： detection accuracy, number of support vectors, false positive rate, falsc negative rate, training time and testing time.
文摘By introducing the distribution of the light energy density in GaN-based light-emitting diode (LED),theLED model based on the incoherent regime and the light extraction efficiency are investigated.The energy density asa function of the angle of incidence is calculated to demonstrate the mechanism of the light extraction.The deviationbetween the tendencies of the transmissivity of the output layer and the extraction efficiency is also demonstrated.
