The ERA-Interim reanalysis wind based on the distance-weighted average remapping for studying the wind circulation in Nigeria is presented. The wind flow using this atmospheric model simulation is studied for identifi...The ERA-Interim reanalysis wind based on the distance-weighted average remapping for studying the wind circulation in Nigeria is presented. The wind flow using this atmospheric model simulation is studied for identification of grid-tie electrification opportunities in different wind locations. A 10-year reanalysis wind speed components at a surface level of the planetary layer at 0.25° × 0.25° spatial resolution is obtained and remapped into a new horizontal wind field at a grid resolution of 0.125° × 0.125° covering longitudinal and latitudinal directions of 3.0 - 15.0°E and 15.0 - 3.0°N, respectively. Using the distance-weighted average technique, the remapped wind field at a new grid resolution of 0.125° × 0.125° is compared at different terrain elevations and approximated close to the actual wind field of the same resolution. To determine the suitability of the prevailing wind for small-scale energy conversion, the magnitude of wind flow across the remapped wind field is studied for a 10-year period. Analysis shows that northern regions of Nigeria have a fair wind potential for a stand-alone application based on the wind flow originated at Gulf of Guinea as well as Chad and Niger. Furthermore, hourly surface wind speed observations from 18 synoptic stations in Nigeria are obtained and compared with the bilinear interpolated wind stations. The reanalysis wind reflects the surface wind observations and proves that the prevailing wind in Nigeria is higher than the reanalysis wind projection obtained from gridded data at resolution of 0.125° × 0.125°. The sectorwise wind directions at each synoptic stations for a period of 10 years are presented.展开更多
SRAM-based computing-in-memory(SRAM-CIM)is expected to solve the“Memory Wall”problem.For the digital domain SRAM-CIM,full-precision digital logic has been utilized to achieve high computational accuracy.However,the ...SRAM-based computing-in-memory(SRAM-CIM)is expected to solve the“Memory Wall”problem.For the digital domain SRAM-CIM,full-precision digital logic has been utilized to achieve high computational accuracy.However,the energy and area efficiency advantages of CIM cannot be fully utilized under error-resilient neural networks(NNs)with given quantization bit-width.Therefore,an all-digital Bit-wise Approximate compressor configurable In-SRAM-computing macro for Energy-efficient NN acceleration,with a data-aware weight Remapping method(BASER),is proposed in this paper.Leveraging the NN error resilience property,six energy-efficient bit-wise compressor configurations are presented under 4b/4b and 3b/3b NN quantization,respectively.Concurrently,a data-aware weight remapping approach is proposed to enhance the NN accuracy without supplementary retraining further.Evaluations of VGG-9 and ResNet-18 on CIFAR-10 and CIFAR-100 datasets show that the proposed BASER achieves 1.35x and 1.29x improvement in energy efficiency,as well as limited accuracy loss and improved NN accuracy,as compared to the previous full-precision and approximate SRAM-CIM design,respectively.展开更多
采用拉氏时间推进加重映到初始网格的方式,在结构化交错欧拉网格上实现一种新型两步欧拉法。拉氏时间推进采用预估-校正方法,混合网格的拉氏计算中引入压力松弛模型。用MOF(Moment of Fluid)重构显式界面,将混合网格剖分为多个介质多面...采用拉氏时间推进加重映到初始网格的方式,在结构化交错欧拉网格上实现一种新型两步欧拉法。拉氏时间推进采用预估-校正方法,混合网格的拉氏计算中引入压力松弛模型。用MOF(Moment of Fluid)重构显式界面,将混合网格剖分为多个介质多面体,实现了精确的相交重映。考虑到已有拉氏网格与拉氏网格相交算法的低效性,实现了与两步欧拉法更适配的拉氏网格与欧拉网格相交算法。数值模拟结果表明:在欧拉框架下构造显式界面,能够提高欧拉方法对界面的分辨能力,本文构造显式界面进行相交重映的算法具有健壮且高效的特点,在大变形模拟中也可以保持较好的完整性。展开更多
Sheet metal spinning is an incremental forming process for producing axisymmetric thinwalled parts through continuous local deformation under the action of rollers.While studying the spinning process by finite element...Sheet metal spinning is an incremental forming process for producing axisymmetric thinwalled parts through continuous local deformation under the action of rollers.While studying the spinning process by finite element(FE)method,a critical bottleneck is the enormous simulation time.For beating off this challenge,a novel multi-mesh method is developed.The method can dynamically track the movement of rollers and adaptively refine the mesh.Thus,a locally refined quadrilateral computation mesh can be generated in the locally-deforming zone and reduce the unnecessary fine elements outside the locally-deforming zone.In the multi-mesh system,the fine elements and coarse elements are extracted from a storage mesh and a background mesh,respectively.Meanwhile,the hanging nodes in the locally refined mesh are removed by designing 4-refinement templates.Between computation mesh and storage mesh,a bi-cubic parametric surface fitting algorithm and accurate remapping methods are conducted to transmit geometric information and physical fields.The proposed method has been verified by two spinning processes.The results suggest that the method can save time by up to about 67%with satisfactory accuracy,especially for distributions of thickness and strain compared with the fully refined mesh.展开更多
It is found that the solution remapping technique proposed in[Numer.Math.Theor.Meth.Appl.,2020,13(4)]and[J.Sci.Comput.,2021,87(3):1-26]does not work out for the Navier-Stokes equations with a high Reynolds number.The ...It is found that the solution remapping technique proposed in[Numer.Math.Theor.Meth.Appl.,2020,13(4)]and[J.Sci.Comput.,2021,87(3):1-26]does not work out for the Navier-Stokes equations with a high Reynolds number.The shape deformations usually reach several boundary layer mesh sizes for viscous flow,which far exceed one-layer mesh that the original method can tolerate.The direct application to Navier-Stokes equations can result in the unphysical pressures in remapped solutions,even though the conservative variables are within the reasonable range.In this work,a new solution remapping technique with lower bound preservation is proposed to construct initial values for the new shapes,and the global minimum density and pressure of the current shape which serve as lower bounds of the corresponding variables are used to constrain the remapped solutions.The solution distribution provided by the present method is proven to be acceptable as an initial value for the new shape.Several numerical experiments show that the present technique can substantially accelerate the flow convergence for large deformation problemswith 70%-80%CPU time reduction in the viscous airfoil drag minimization.展开更多
文摘The ERA-Interim reanalysis wind based on the distance-weighted average remapping for studying the wind circulation in Nigeria is presented. The wind flow using this atmospheric model simulation is studied for identification of grid-tie electrification opportunities in different wind locations. A 10-year reanalysis wind speed components at a surface level of the planetary layer at 0.25° × 0.25° spatial resolution is obtained and remapped into a new horizontal wind field at a grid resolution of 0.125° × 0.125° covering longitudinal and latitudinal directions of 3.0 - 15.0°E and 15.0 - 3.0°N, respectively. Using the distance-weighted average technique, the remapped wind field at a new grid resolution of 0.125° × 0.125° is compared at different terrain elevations and approximated close to the actual wind field of the same resolution. To determine the suitability of the prevailing wind for small-scale energy conversion, the magnitude of wind flow across the remapped wind field is studied for a 10-year period. Analysis shows that northern regions of Nigeria have a fair wind potential for a stand-alone application based on the wind flow originated at Gulf of Guinea as well as Chad and Niger. Furthermore, hourly surface wind speed observations from 18 synoptic stations in Nigeria are obtained and compared with the bilinear interpolated wind stations. The reanalysis wind reflects the surface wind observations and proves that the prevailing wind in Nigeria is higher than the reanalysis wind projection obtained from gridded data at resolution of 0.125° × 0.125°. The sectorwise wind directions at each synoptic stations for a period of 10 years are presented.
基金supported in part by the National Key R&D Program of China under Grant 2023YFB450220in part by the National Natural Science Foundation of China under Grant 62174110 and Grant 62104025in part by the Natural Science Foundation of Shanghai under Grant 23ZR1433200.
文摘SRAM-based computing-in-memory(SRAM-CIM)is expected to solve the“Memory Wall”problem.For the digital domain SRAM-CIM,full-precision digital logic has been utilized to achieve high computational accuracy.However,the energy and area efficiency advantages of CIM cannot be fully utilized under error-resilient neural networks(NNs)with given quantization bit-width.Therefore,an all-digital Bit-wise Approximate compressor configurable In-SRAM-computing macro for Energy-efficient NN acceleration,with a data-aware weight Remapping method(BASER),is proposed in this paper.Leveraging the NN error resilience property,six energy-efficient bit-wise compressor configurations are presented under 4b/4b and 3b/3b NN quantization,respectively.Concurrently,a data-aware weight remapping approach is proposed to enhance the NN accuracy without supplementary retraining further.Evaluations of VGG-9 and ResNet-18 on CIFAR-10 and CIFAR-100 datasets show that the proposed BASER achieves 1.35x and 1.29x improvement in energy efficiency,as well as limited accuracy loss and improved NN accuracy,as compared to the previous full-precision and approximate SRAM-CIM design,respectively.
文摘采用拉氏时间推进加重映到初始网格的方式,在结构化交错欧拉网格上实现一种新型两步欧拉法。拉氏时间推进采用预估-校正方法,混合网格的拉氏计算中引入压力松弛模型。用MOF(Moment of Fluid)重构显式界面,将混合网格剖分为多个介质多面体,实现了精确的相交重映。考虑到已有拉氏网格与拉氏网格相交算法的低效性,实现了与两步欧拉法更适配的拉氏网格与欧拉网格相交算法。数值模拟结果表明:在欧拉框架下构造显式界面,能够提高欧拉方法对界面的分辨能力,本文构造显式界面进行相交重映的算法具有健壮且高效的特点,在大变形模拟中也可以保持较好的完整性。
基金co-supported by the supports of Guangdong Basic and Applied Basic Research Foundation(No.2019B1515120047)the National Natural Science Foundation of China(No.52130507)。
文摘Sheet metal spinning is an incremental forming process for producing axisymmetric thinwalled parts through continuous local deformation under the action of rollers.While studying the spinning process by finite element(FE)method,a critical bottleneck is the enormous simulation time.For beating off this challenge,a novel multi-mesh method is developed.The method can dynamically track the movement of rollers and adaptively refine the mesh.Thus,a locally refined quadrilateral computation mesh can be generated in the locally-deforming zone and reduce the unnecessary fine elements outside the locally-deforming zone.In the multi-mesh system,the fine elements and coarse elements are extracted from a storage mesh and a background mesh,respectively.Meanwhile,the hanging nodes in the locally refined mesh are removed by designing 4-refinement templates.Between computation mesh and storage mesh,a bi-cubic parametric surface fitting algorithm and accurate remapping methods are conducted to transmit geometric information and physical fields.The proposed method has been verified by two spinning processes.The results suggest that the method can save time by up to about 67%with satisfactory accuracy,especially for distributions of thickness and strain compared with the fully refined mesh.
基金This project is supported by the National Natural Science Foundation of China(No.12001031).
文摘It is found that the solution remapping technique proposed in[Numer.Math.Theor.Meth.Appl.,2020,13(4)]and[J.Sci.Comput.,2021,87(3):1-26]does not work out for the Navier-Stokes equations with a high Reynolds number.The shape deformations usually reach several boundary layer mesh sizes for viscous flow,which far exceed one-layer mesh that the original method can tolerate.The direct application to Navier-Stokes equations can result in the unphysical pressures in remapped solutions,even though the conservative variables are within the reasonable range.In this work,a new solution remapping technique with lower bound preservation is proposed to construct initial values for the new shapes,and the global minimum density and pressure of the current shape which serve as lower bounds of the corresponding variables are used to constrain the remapped solutions.The solution distribution provided by the present method is proven to be acceptable as an initial value for the new shape.Several numerical experiments show that the present technique can substantially accelerate the flow convergence for large deformation problemswith 70%-80%CPU time reduction in the viscous airfoil drag minimization.
