Estimated daily flows along the Yarlung Zangbo River(YZR)for periods well before stream gauging began in the 1950s are needed to characterize long-term flow variabilities.To examine whether the recorded flow in recent...Estimated daily flows along the Yarlung Zangbo River(YZR)for periods well before stream gauging began in the 1950s are needed to characterize long-term flow variabilities.To examine whether the recorded flow in recent decades can sufficiently represent natural variability and change,daily flows were reconstructed using the autocalibrated variable infiltration capacity(VIC)model using preinstrumental proxy-based precipitation and temperature data for four mainstream stations in the YZR basin for the period 1473–2000.VIC performed well during calibration and validation for all stations,with the Nash–Sutcliffe coefficient of efficiency exceeding 0.55 and the correlation coefficient and the bias in the standard deviation being close to 1.This study showed that the annual or decadal flows have not significantly changed over the past 528 years.The longterm mean flow differed from that for the period 1951–2000 by only 6%–8%,although the reconstructed maximum annual flow was approximately 24%–35%higher and the minimum flow was 0%–9%lower than those for 1951−2000.Moreover,the flow in recent decades could not capture the persistently high or low flows and some extreme wet/dry years found in the preinstrumental period.The timing of the maximum daily flow was more concentrated,and its magnitude was slightly but not significantly higher.A comparison of flows at the four mainstream stations indicated that the annual and peak daily flows were not synchronous.The timing of floods was highly correlated among the three stations,apart from that at Lazi.Furthermore,flows at Nuxia demonstrated a pronounced response to the Indian Ocean Dipole approximately at 20-year periodicity,whereas those at Lazi were more strongly influenced by the North Atlantic Oscillation on an~50-year scale.This study improves our understanding of the long-term flow variability for the whole YZR basin and its potential climatic drivers for better water resource management and more effective flood hazard control.展开更多
The internal flow fields within a three-dimensional inward-tunning combined inlet are extremely complex,especially during the engine mode transition,where the tunnel changes may impact the flow fields significantly.To...The internal flow fields within a three-dimensional inward-tunning combined inlet are extremely complex,especially during the engine mode transition,where the tunnel changes may impact the flow fields significantly.To develop an efficient flow field reconstruction model for this,we present an Improved Conditional Denoising Diffusion Generative Adversarial Network(ICDDGAN),which integrates Conditional Denoising Diffusion Probabilistic Models(CDDPMs)with Style GAN,and introduce a reconstruction discrimination mechanism and dynamic loss weight learning strategy.We establish the Mach number flow field dataset by numerical simulation at various backpressures for the mode transition process from turbine mode to ejector ramjet mode at Mach number 2.5.The proposed ICDDGAN model,given only sparse parameter information,can rapidly generate high-quality Mach number flow fields without a large number of samples for training.The results show that ICDDGAN is superior to CDDGAN in terms of training convergence and stability.Moreover,the interpolation and extrapolation test results during backpressure conditions show that ICDDGAN can accurately and quickly reconstruct Mach number fields at various tunnel slice shapes,with a Structural Similarity Index Measure(SSIM)of over 0.96 and a Mean-Square Error(MSE)of 0.035%to actual flow fields,reducing time costs by 7-8 orders of magnitude compared to Computational Fluid Dynamics(CFD)calculations.This can provide an efficient means for rapid computation of complex flow fields.展开更多
This study presented a hybrid model method based on proper orthogonal decomposition(POD) for flow field reconstructions and aerodynamic design optimization. The POD basis modes have better description performance in a...This study presented a hybrid model method based on proper orthogonal decomposition(POD) for flow field reconstructions and aerodynamic design optimization. The POD basis modes have better description performance in a system space compared to the widely used semi-empirical basis functions because they are obtained through singular value decomposition of the system.Instead of the widely used linear regression, nonlinear regression methods are used in the function response of the coefficients of POD basis modes. Moreover, an adaptive Latin hypercube design method with improved space filling and correlation based on a multi-objective optimization approach was employed to supply the necessary samples. Prior to design optimization, the response performance of POD-based hybrid models was first investigated and validated through flow reconstructions of both single-and multiple blade rows. Then, an inverse design was performed to approach a given spanwise flow turning distribution at the outlet of a turbine blade by changing the spanwise stagger angle, based on the hybrid model method. Finally, the span wise blade sweep of a transonic compressor rotor and the spanwise stagger angle of the stator blade of a single low-speed compressor stage were modified to reduce the flow losses with the constraints of mass flow rate, total pressure ratio, and outlet flow turning.The results are presented in detail, demonstrating the good response performance of POD-based hybrid models on missing data reconstructions and the effectiveness of POD-based hybrid model method in aerodynamic design optimization.展开更多
The transient cavitating flow around the Clark-Y hydrofoil is numerically investigated by the dynamic mode decomposition with criterion.Based on the ranking dominant modes,frequencies of the first four modes are in go...The transient cavitating flow around the Clark-Y hydrofoil is numerically investigated by the dynamic mode decomposition with criterion.Based on the ranking dominant modes,frequencies of the first four modes are in good accordance with those obtained by fast Fourier transform.Furthermore,the cavitating flow field is reconstructed by the first four modes,and the dominant flow features are well captured with the reconstructed error below 12%when compared to the simulated flow field.This paper offers a reference for observing and reconstructing the flow fields,and gives a novel insight into the transient cavitating flow features.展开更多
In this paper, on the basis of experimental data of two kinds of chemical explosions, the piston-pushing model of spherical blast-waves and the second-order Godunov-type scheme of finite difference methods with high i...In this paper, on the basis of experimental data of two kinds of chemical explosions, the piston-pushing model of spherical blast-waves and the second-order Godunov-type scheme of finite difference methods with high identification to discontinuity are used to the numerical reconstruction of part of an actual hemispherical blast-wave flow field by properly adjusting the moving bounary conditions of a piston. This method is simple and reliable. It is suitable to the evaluation of effects of the blast-wave flow field away from the explosion center.展开更多
Flow field computation plays a critical role in both scientific research and engineering applications.For decades,computational fluid dynamics(CFD)has served as the cornerstone of flow field analysis;however,high-reso...Flow field computation plays a critical role in both scientific research and engineering applications.For decades,computational fluid dynamics(CFD)has served as the cornerstone of flow field analysis;however,high-resolution simulations are often hindered by considerable computational costs and lengthy processing times.In recent years,deep learning(DL),with its exceptional capability to handle high-dimensional nonlinear problems,has achieved remarkable progress in the field of fluid mechanics.This paper provides a comprehensive review of recent advances in applying DL methods to accelerate flow field computation,with particular emphasis on complex indoor environments characterized by multi-physics coupling.We begin by outlining the fundamental frameworks of deep learning and,on this basis,summarize four representative neural network architectures for flow prediction:end-to-end mapping networks,reduced-order mapping networks,physics-informed neural networks(PINNs),and operator neural networks(ONNs).We then systematically review the specific applications of these DL algorithms in indoor flow field prediction.In addition,we discuss key challenges faced by current research,including the lack of large,high-quality databases,limited interpretability and generalization capability of existing models,and the difficulty of accurately representing real indoor environments.Finally,we propose several promising research directions,such as exploring advanced algorithms,enhancing self-supervised learning techniques,and developing geometry-aware network models and multi-task hybrid frameworks.Advancing these frontiers is expected not only to significantly improve the efficiency and accuracy of flow field computations but also to provide a solid theoretical foundation and technical support for the optimization of intelligent indoor environments.展开更多
基金supported by the National Natural Science Foundation of China(Grant Nos.91747207,U2340223)the Second Tibetan Plateau Scientific Expedition and Research Program(STEP)(Grant No.2019QZKK0903).
文摘Estimated daily flows along the Yarlung Zangbo River(YZR)for periods well before stream gauging began in the 1950s are needed to characterize long-term flow variabilities.To examine whether the recorded flow in recent decades can sufficiently represent natural variability and change,daily flows were reconstructed using the autocalibrated variable infiltration capacity(VIC)model using preinstrumental proxy-based precipitation and temperature data for four mainstream stations in the YZR basin for the period 1473–2000.VIC performed well during calibration and validation for all stations,with the Nash–Sutcliffe coefficient of efficiency exceeding 0.55 and the correlation coefficient and the bias in the standard deviation being close to 1.This study showed that the annual or decadal flows have not significantly changed over the past 528 years.The longterm mean flow differed from that for the period 1951–2000 by only 6%–8%,although the reconstructed maximum annual flow was approximately 24%–35%higher and the minimum flow was 0%–9%lower than those for 1951−2000.Moreover,the flow in recent decades could not capture the persistently high or low flows and some extreme wet/dry years found in the preinstrumental period.The timing of the maximum daily flow was more concentrated,and its magnitude was slightly but not significantly higher.A comparison of flows at the four mainstream stations indicated that the annual and peak daily flows were not synchronous.The timing of floods was highly correlated among the three stations,apart from that at Lazi.Furthermore,flows at Nuxia demonstrated a pronounced response to the Indian Ocean Dipole approximately at 20-year periodicity,whereas those at Lazi were more strongly influenced by the North Atlantic Oscillation on an~50-year scale.This study improves our understanding of the long-term flow variability for the whole YZR basin and its potential climatic drivers for better water resource management and more effective flood hazard control.
文摘The internal flow fields within a three-dimensional inward-tunning combined inlet are extremely complex,especially during the engine mode transition,where the tunnel changes may impact the flow fields significantly.To develop an efficient flow field reconstruction model for this,we present an Improved Conditional Denoising Diffusion Generative Adversarial Network(ICDDGAN),which integrates Conditional Denoising Diffusion Probabilistic Models(CDDPMs)with Style GAN,and introduce a reconstruction discrimination mechanism and dynamic loss weight learning strategy.We establish the Mach number flow field dataset by numerical simulation at various backpressures for the mode transition process from turbine mode to ejector ramjet mode at Mach number 2.5.The proposed ICDDGAN model,given only sparse parameter information,can rapidly generate high-quality Mach number flow fields without a large number of samples for training.The results show that ICDDGAN is superior to CDDGAN in terms of training convergence and stability.Moreover,the interpolation and extrapolation test results during backpressure conditions show that ICDDGAN can accurately and quickly reconstruct Mach number fields at various tunnel slice shapes,with a Structural Similarity Index Measure(SSIM)of over 0.96 and a Mean-Square Error(MSE)of 0.035%to actual flow fields,reducing time costs by 7-8 orders of magnitude compared to Computational Fluid Dynamics(CFD)calculations.This can provide an efficient means for rapid computation of complex flow fields.
基金supported by the National Natural Science Foundation of China(Grant Nos.51676003,51206003 and 51376009)
文摘This study presented a hybrid model method based on proper orthogonal decomposition(POD) for flow field reconstructions and aerodynamic design optimization. The POD basis modes have better description performance in a system space compared to the widely used semi-empirical basis functions because they are obtained through singular value decomposition of the system.Instead of the widely used linear regression, nonlinear regression methods are used in the function response of the coefficients of POD basis modes. Moreover, an adaptive Latin hypercube design method with improved space filling and correlation based on a multi-objective optimization approach was employed to supply the necessary samples. Prior to design optimization, the response performance of POD-based hybrid models was first investigated and validated through flow reconstructions of both single-and multiple blade rows. Then, an inverse design was performed to approach a given spanwise flow turning distribution at the outlet of a turbine blade by changing the spanwise stagger angle, based on the hybrid model method. Finally, the span wise blade sweep of a transonic compressor rotor and the spanwise stagger angle of the stator blade of a single low-speed compressor stage were modified to reduce the flow losses with the constraints of mass flow rate, total pressure ratio, and outlet flow turning.The results are presented in detail, demonstrating the good response performance of POD-based hybrid models on missing data reconstructions and the effectiveness of POD-based hybrid model method in aerodynamic design optimization.
基金the National Key R&D Program of China(Grants 2016YFC0300800 and 2016YFC0300802)the National Natural Science Foundation of China(Grants 11772340 and 11672315)the Science and Technology on Water Jet Propulsion Laboratory(Grant 6142223190101).
文摘The transient cavitating flow around the Clark-Y hydrofoil is numerically investigated by the dynamic mode decomposition with criterion.Based on the ranking dominant modes,frequencies of the first four modes are in good accordance with those obtained by fast Fourier transform.Furthermore,the cavitating flow field is reconstructed by the first four modes,and the dominant flow features are well captured with the reconstructed error below 12%when compared to the simulated flow field.This paper offers a reference for observing and reconstructing the flow fields,and gives a novel insight into the transient cavitating flow features.
文摘In this paper, on the basis of experimental data of two kinds of chemical explosions, the piston-pushing model of spherical blast-waves and the second-order Godunov-type scheme of finite difference methods with high identification to discontinuity are used to the numerical reconstruction of part of an actual hemispherical blast-wave flow field by properly adjusting the moving bounary conditions of a piston. This method is simple and reliable. It is suitable to the evaluation of effects of the blast-wave flow field away from the explosion center.
基金This work was supported by the National Natural Science Foundation of China(Grant No.52006232)the Youth Innovation Promotion Association of Chinese Academy of Sciences(Grant No.2019020)。
基金supported by the National Natural Science Foundation of China(grant number 52178072).
文摘Flow field computation plays a critical role in both scientific research and engineering applications.For decades,computational fluid dynamics(CFD)has served as the cornerstone of flow field analysis;however,high-resolution simulations are often hindered by considerable computational costs and lengthy processing times.In recent years,deep learning(DL),with its exceptional capability to handle high-dimensional nonlinear problems,has achieved remarkable progress in the field of fluid mechanics.This paper provides a comprehensive review of recent advances in applying DL methods to accelerate flow field computation,with particular emphasis on complex indoor environments characterized by multi-physics coupling.We begin by outlining the fundamental frameworks of deep learning and,on this basis,summarize four representative neural network architectures for flow prediction:end-to-end mapping networks,reduced-order mapping networks,physics-informed neural networks(PINNs),and operator neural networks(ONNs).We then systematically review the specific applications of these DL algorithms in indoor flow field prediction.In addition,we discuss key challenges faced by current research,including the lack of large,high-quality databases,limited interpretability and generalization capability of existing models,and the difficulty of accurately representing real indoor environments.Finally,we propose several promising research directions,such as exploring advanced algorithms,enhancing self-supervised learning techniques,and developing geometry-aware network models and multi-task hybrid frameworks.Advancing these frontiers is expected not only to significantly improve the efficiency and accuracy of flow field computations but also to provide a solid theoretical foundation and technical support for the optimization of intelligent indoor environments.
