The minimum-time path for intercepting a moving target with a prescribed impact angle is studied in the paper.The candidate paths from Pontryagin’s maximum principle are synthesized,so that each candidate is related ...The minimum-time path for intercepting a moving target with a prescribed impact angle is studied in the paper.The candidate paths from Pontryagin’s maximum principle are synthesized,so that each candidate is related to a zero of a real-valued function.It is found that the real-valued functions or their first-order derivatives can be converted to polynomials of at most fourth degree.As a result,each candidate path can be computed within a constant time by embedding a standard polynomial solver into the typical bisection method.The control strategy along the shortest candidate eventually gives rise to the time-optimal guidance law.Finally,the developments of the paper are illustrated and verified by three numerical examples.展开更多
Variable-sweep wings have large shape-changing capabilities and wide flight envelops,which are considered as one of the most promising directions for intelligent morphing UAVs.Aerodynamic investigations always focus o...Variable-sweep wings have large shape-changing capabilities and wide flight envelops,which are considered as one of the most promising directions for intelligent morphing UAVs.Aerodynamic investigations always focus on several static states in the varying sweep process,which ignore the unsteady aerodynamic characteristics.However,deviations to static aerodynamic forces are inevitably caused by dynamic sweep motion.In this work,first,unsteady aerodynamic characteristics on a typical variable-sweep UAV with large aspect ratio were analyzed.Then,deep mechanism of unsteady aerodynamic characteristics in the varying sweep process was studied.Finally,numerical simulation method integrated with structured moving overset grids was applied to solve the unsteady fluid of varying sweep process.The simulation results of a sweep forward-backward circle show a distinct dynamic hysteresis loop surrounding the static data for the aerodynamic forces.Compared with the static lift coefficients,at the same sweep angles,dynamic lift coefficient in sweep forward process are all smaller,while dynamic sweep backward lift coefficient are all larger.In addition,dynamic deviations to static lift coefficient are positively related with the varying sweep speeds.Mechanism study on the unsteady aerodynamic characteristics indicates that three key factors lead to the dynamic hysteresis loop in varying sweep process.They are the effects of additional velocity caused by varying sweep motion,the effects of flow hysteresis and viscosity.The additional velocity induced by sweep motion affects the transversal flow direction along the wing and the effective angle of attack at the airfoil profile.The physical properties of flow,the hysteresis and viscosity affect the unsteady aerodynamic characteristics by flow separation and induced vortexes.展开更多
A direct-forcing fictitious domain(DFFD) method is used to perform fully resolved numerical simulations of turbulent channel flows laden with large neutrally buoyant particles. The effects of the particles on the turb...A direct-forcing fictitious domain(DFFD) method is used to perform fully resolved numerical simulations of turbulent channel flows laden with large neutrally buoyant particles. The effects of the particles on the turbulence(including the mean velocity,the root mean square(RMS) of the velocity fluctuation, the probability density function(PDF) of the velocity, and the vortex structures) at a friction Reynolds number of 395 are investigated. The results show that the drag-reduction effect caused by finite-size spherical particles at low particle volumes is negligibly small. The particle effects on the RMS velocities at Re_τ = 395 are significantly smaller than those at Re_τ = 180, despite qualitatively the same effects, i.e., the presence of particles decreases the maximum streamwise RMS velocity near the wall via weakening the large-scale streamwise vortices,and increases the transverse and spanwise RMS velocities in the vicinity of the wall by inducing smaller-scale vortices. The effects of the particles on the PDFs of the fluid fluctuating velocities normalized with the RMS velocities are small, regardless of the particle size, the particle volume fraction, and the Reynolds number.展开更多
Physical monotonicity is a pervasive phenomenon in the aerodynamic characteristics of aircraft,where the aerodynamic lift consistently increases with the angle of attack within the stalling range.Existing machine lear...Physical monotonicity is a pervasive phenomenon in the aerodynamic characteristics of aircraft,where the aerodynamic lift consistently increases with the angle of attack within the stalling range.Existing machine learning models for aerodynamic predic-tions often overlook this monotonicity,resulting in poor interpretability and credibility.To address this issue,we introduce a monotonic model,the Deep Lattice Network,which integrates the monotonicity constraint of the lift coefficient into machine learn-ing based aerodynamic prediction framework.In this paper,we propose a novel deep learning model,Deep Lattice Cross Network,which aims to rapidly predict aerody-namic forces with high precision while ensuring monotonic constraints.Multi-Task Learning method is utilized to simultaneously predict both lift and drag coefficients,thereby enhancing the efficiency of the model.To optimize the training process and minimize costs,we adopt a unique two-phase deep network training strategy.Based on computational fluid dynamics simulation datasets of a morphing aircraft,the model is trained,and the efficacy of the model is tested by two interpolation and two extrapolation datasets.The results show a remarkable alignment with com-putational fluid dynamics outcomes across all test scenarios.Extended testing across a wider range of attack angles further highlights the superiority of the Deep Lat-tice Cross Network in upholding monotonicity.Incorporating monotonicity constraints not only improves predictive accuracy of the model but also greatly enhances its physi-cal interpretability,which is crucial for advancing the development of more depend-able aerodynamic prediction models.展开更多
基金supported by the National Natural Science Foundation of China(Nos.61903331,62088101)the Shanghai Aerospace Science and Technology Innovation Fund,China(No.SAST2019-10)。
文摘The minimum-time path for intercepting a moving target with a prescribed impact angle is studied in the paper.The candidate paths from Pontryagin’s maximum principle are synthesized,so that each candidate is related to a zero of a real-valued function.It is found that the real-valued functions or their first-order derivatives can be converted to polynomials of at most fourth degree.As a result,each candidate path can be computed within a constant time by embedding a standard polynomial solver into the typical bisection method.The control strategy along the shortest candidate eventually gives rise to the time-optimal guidance law.Finally,the developments of the paper are illustrated and verified by three numerical examples.
基金supported by the National Natural Science Foundation of China(No.12202384)the Rotor Aerodynamics Key Laboratory Foundation of China Aerodynamics Research and Development Center(No.2108RAL202102-5).
文摘Variable-sweep wings have large shape-changing capabilities and wide flight envelops,which are considered as one of the most promising directions for intelligent morphing UAVs.Aerodynamic investigations always focus on several static states in the varying sweep process,which ignore the unsteady aerodynamic characteristics.However,deviations to static aerodynamic forces are inevitably caused by dynamic sweep motion.In this work,first,unsteady aerodynamic characteristics on a typical variable-sweep UAV with large aspect ratio were analyzed.Then,deep mechanism of unsteady aerodynamic characteristics in the varying sweep process was studied.Finally,numerical simulation method integrated with structured moving overset grids was applied to solve the unsteady fluid of varying sweep process.The simulation results of a sweep forward-backward circle show a distinct dynamic hysteresis loop surrounding the static data for the aerodynamic forces.Compared with the static lift coefficients,at the same sweep angles,dynamic lift coefficient in sweep forward process are all smaller,while dynamic sweep backward lift coefficient are all larger.In addition,dynamic deviations to static lift coefficient are positively related with the varying sweep speeds.Mechanism study on the unsteady aerodynamic characteristics indicates that three key factors lead to the dynamic hysteresis loop in varying sweep process.They are the effects of additional velocity caused by varying sweep motion,the effects of flow hysteresis and viscosity.The additional velocity induced by sweep motion affects the transversal flow direction along the wing and the effective angle of attack at the airfoil profile.The physical properties of flow,the hysteresis and viscosity affect the unsteady aerodynamic characteristics by flow separation and induced vortexes.
基金Project supported by the National Natural Science Foundation of China(Nos.91752117,11632016,and 11372275)the Natural Science Foundation of Zhejiang Province of China(No.LY17A020005)
文摘A direct-forcing fictitious domain(DFFD) method is used to perform fully resolved numerical simulations of turbulent channel flows laden with large neutrally buoyant particles. The effects of the particles on the turbulence(including the mean velocity,the root mean square(RMS) of the velocity fluctuation, the probability density function(PDF) of the velocity, and the vortex structures) at a friction Reynolds number of 395 are investigated. The results show that the drag-reduction effect caused by finite-size spherical particles at low particle volumes is negligibly small. The particle effects on the RMS velocities at Re_τ = 395 are significantly smaller than those at Re_τ = 180, despite qualitatively the same effects, i.e., the presence of particles decreases the maximum streamwise RMS velocity near the wall via weakening the large-scale streamwise vortices,and increases the transverse and spanwise RMS velocities in the vicinity of the wall by inducing smaller-scale vortices. The effects of the particles on the PDFs of the fluid fluctuating velocities normalized with the RMS velocities are small, regardless of the particle size, the particle volume fraction, and the Reynolds number.
基金supported by the National Natural Science Foundation of China(Grants No.12202384 and No.U2241274)the Defense Industrial Technology Development Program(Grants No.JCKY2023205B013 and No.JCKY2021205B003).
文摘Physical monotonicity is a pervasive phenomenon in the aerodynamic characteristics of aircraft,where the aerodynamic lift consistently increases with the angle of attack within the stalling range.Existing machine learning models for aerodynamic predic-tions often overlook this monotonicity,resulting in poor interpretability and credibility.To address this issue,we introduce a monotonic model,the Deep Lattice Network,which integrates the monotonicity constraint of the lift coefficient into machine learn-ing based aerodynamic prediction framework.In this paper,we propose a novel deep learning model,Deep Lattice Cross Network,which aims to rapidly predict aerody-namic forces with high precision while ensuring monotonic constraints.Multi-Task Learning method is utilized to simultaneously predict both lift and drag coefficients,thereby enhancing the efficiency of the model.To optimize the training process and minimize costs,we adopt a unique two-phase deep network training strategy.Based on computational fluid dynamics simulation datasets of a morphing aircraft,the model is trained,and the efficacy of the model is tested by two interpolation and two extrapolation datasets.The results show a remarkable alignment with com-putational fluid dynamics outcomes across all test scenarios.Extended testing across a wider range of attack angles further highlights the superiority of the Deep Lat-tice Cross Network in upholding monotonicity.Incorporating monotonicity constraints not only improves predictive accuracy of the model but also greatly enhances its physi-cal interpretability,which is crucial for advancing the development of more depend-able aerodynamic prediction models.
