In this paper,the adaptive fault-tolerant formation tracking control problem for a set of heterogeneous unmanned aerial vehicle(UAV)and unmanned ground vehicle(UGV)systems with actuator loss of effectiveness faults is...In this paper,the adaptive fault-tolerant formation tracking control problem for a set of heterogeneous unmanned aerial vehicle(UAV)and unmanned ground vehicle(UGV)systems with actuator loss of effectiveness faults is investigated.The cooperative fault-tolerant formation control strategy for UAV and UGV collaborative systems is classi-ed into the altitude consensus control scheme for follower UAVs and the position cooperative formation control scheme for all followers.The altitude consensus control algorithm is designed by utilizing backstepping control technique to drive all UAVs to a desired prede-ned height.Then,based on synchronization formation error information,the position cooperative formation control algorithm is proposed for all followers to reach the expected position and perform the desired formation con-guration.The adaptive fault estimation term is adopted in the designed fault-tolerant formation control algorithm to compensate for the actuator loss of effectiveness fault.Finally,a simulation example is proposed to reveal the validity of the designed cooperative formation tracking control scheme.展开更多
In this mini-review we summarize the progress of modeling, simulation and analysis of shock responses of heterogeneous materials in our group in recent years. The basic methodology is as below. We first decompose the ...In this mini-review we summarize the progress of modeling, simulation and analysis of shock responses of heterogeneous materials in our group in recent years. The basic methodology is as below. We first decompose the problem into different scales. Construct/Choose a model according to the scale and main mechanisms working at that scale. Perform numerical simulations using the relatively mature schemes. The physical information is transferred between neighboring scales in such a way: The statistical information of results in smaller scale contributes to establishing the constitutive equation in larger one. Except for the microscopic Molecular Dynamics(MD) model, both the mesoscopic and macroscopic models can be further classified into two categories,solidic and fluidic models, respectively. The basic ideas and key techniques of the MD, material point method and discrete Boltzmann method are briefly reviewed. Among various schemes used in analyzing the complex fields and structures, the morphological analysis and the home-built software, GISO, are briefly introduced. New observations are summarized for scales from the larger to the smaller.展开更多
Water vapor nucleation on particle's surface plays an important role in dust removal,cloud formation,and particle measurement.However,the selectivity of nucleation sites and the nucleation characteristic of water ...Water vapor nucleation on particle's surface plays an important role in dust removal,cloud formation,and particle measurement.However,the selectivity of nucleation sites and the nucleation characteristic of water molecule on the particle's surface are still unclear,especially for the aggregated particles.In this paper,the effects of particle wettability and aggregation modes on the selectivity of nucleation sites and the nucleation characteristics were investigated using molecular dynamics simulation.The results were compared with our earlier experimental findings.It illustrates how the contact angle of clusters,the growth velocity,and the growth duration are all influenced by the interaction coefficient between water and particles.Moreover,the nucleation sites of water molecules on the particle aggregation surface exhibit a definite selectivity.The primary indicator of this selectivity is the preferential nucleation of water molecules at the interfaces of linear chain aggregation particles,at the inner side of non-linear chain aggregation particles,and at the centers of ring aggregation.These results are in good agreement with our previous experimental findings.More significantly,additional research has revealed that subcritical-size clusters typically aggregate on two-particle surfaces spacing when the spacing smaller than the critical cluster size.展开更多
A two-stage transition upon crossing the glass transition of polystyrene with increasing temperature was precisely determined and interpreted by using solid-state nuclear magnetic resonance （SSNMR）, 1H-XH dipolar co...A two-stage transition upon crossing the glass transition of polystyrene with increasing temperature was precisely determined and interpreted by using solid-state nuclear magnetic resonance （SSNMR）, 1H-XH dipolar couplings based double quantum-filtered （DQF） and dipolar filter （DF） experiments and 13C chemical shift anisotropy （CSA） based centerband-only detection of exchange （CODEX） experiment are used to fully characterize the time scale of molecular motions during the glass transition. While differential scanning calorimetry （DSC） and CODEX experiment predicted the first stage of glass transiton, DQF and DF experiments provided the evidence for the second stage transition during which the time scale of molecular motions changed from very slow （t 〉 ms） to very fast （t 〈 Its）. The first stage of glass transition begins with the occurrence of remarkable slow re-orientation motions of the polymer backbone segments and ends when the degree of slow motion reaches maximum. The onset and endpoint of the conventional calorimetric glass transition of polystyrene can be quantitatively determined at the molecular level by SSNMR. In the second stage, a subsequent dramatic transition associated with the melting of the glassy components was observed. In this stage liquid-like NMR signals appeared and rapidly increased in intensity after a characteristic temperature Tf （-1.1Tg）. The signals associated with the glassy components completely disappeared at another characteristic temperature Tc （-1.2Tg）.展开更多
Much progress has been achieved for both experimental and theoretical studies on the dissociative chemisorption of molecules on surfaces.Quantum state-resolved experimental data has provided unprecedented details for ...Much progress has been achieved for both experimental and theoretical studies on the dissociative chemisorption of molecules on surfaces.Quantum state-resolved experimental data has provided unprecedented details for these fundamental steps in heterogeneous catalysis,while the quantitative dynamics is still not fully understood in theory.An in-depth understanding of experimental observations relies on accurate dynamical calculations,in which the potential energy surface and adequate quantum mechanical implementation are desired.This article summarizes the current methodologies on the construction of potential energy surfaces and the quantum mechanical treatments,some of which are promising for future applications.The challenges in this field are also addressed.展开更多
In opinion dynamics,the convergence of the heterogeneous Hegselmann-Krause(HK) dynamics has always been an open problem for years which looks forward to any essential progress.In this short note,we prove a partial con...In opinion dynamics,the convergence of the heterogeneous Hegselmann-Krause(HK) dynamics has always been an open problem for years which looks forward to any essential progress.In this short note,we prove a partial convergence conclusion of the general heterogeneous HK dynamics.That is,there must be some agents who will reach static states in finite time,while the other opinions have to evolve between them with a minimum distance if all the opinions does not reach consensus.And this result leads to the convergence of several special cases of heterogeneous HK dynamics,including when the minimum confidence bound is large enough,the initial opinion difference is small enough,and so on.展开更多
The results of molecular-dynamics (MD) simulation are obtained for structural and thermodynamic properties of the molten system, Na1-xPbx (x ≤ 0.1), at 698 K in the model of nearly free electronic gas (NFE approximat...The results of molecular-dynamics (MD) simulation are obtained for structural and thermodynamic properties of the molten system, Na1-xPbx (x ≤ 0.1), at 698 K in the model of nearly free electronic gas (NFE approximation). The all numerical experiments are carried out by unified procedure: 1) equilibrating the MD cell 1 ps;2) calculating partial radial distribution functions gab(r) in 1, 5, and 10 ps;3) data handling for calculating other characteristics. It follows from this that lead impurity in liquid sodium at concentration in the range of 1% - 10% is characterized by micro-heterogenetic structure in the form of cluster compounds with variable composition. Just therefore the sodium alloys, Na-MIV, with four-group elements of Periodic table have no eutectic in this range of additive concentrations. This eutectic is needed for modifying sodium coolant of the fast nuclear reactor. Therefore it is reasonable to find an alternative alloy of sodium with additive from adjacent groups which has a eutectic in this range of concentrations, for example, the eutectic, Na0.929Tl0.071, with melting point of 64°C. The modified sodium coolant by isotope, 205Tl, can appear attractive for inhibiting the chemical activity of sodium just as the lead alloyed one.展开更多
Heterogeneous systems with both Central Processing Units （CPUs） and Graphics Processing Units （GPUs） are frequently used to accelerate short-ranged Molecular Dynamics （MD） simulations. The most time-consuming ta...Heterogeneous systems with both Central Processing Units （CPUs） and Graphics Processing Units （GPUs） are frequently used to accelerate short-ranged Molecular Dynamics （MD） simulations. The most time-consuming task in short-ranged MD simulations is the computation of particle-to-particle interac- tions. Beyond a certain distance, these interactions decrease to zero. To minimize the operations to investi- gate distance, previous works have tiled interactions by employing the spatial attribute, which increases the memory access and GPU computations, hence decreasing performance. Other studies ignore the spatial attribute and construct an all-versus-all interaction matrix, which has poor scalability. This paper presents an improved algorithm. The algorithm first bins particles into voxels according to the spatial attributes, and then tiles the all-versus-all matrix into voxel-versus-voxel sub-matrixes. Only the sub-matrixes between neighbor- ing voxels are computed on the GPU. Therefore, the algorithm reduces the distance examine operations and limits additional memory access and GPU computations. This paper also adopts a multi-level program- ming model to implement the algorithm on multi-nodes of Tianhe-lA. By employing （1） a patch design to ex- ploit parallelism across the simulation domain, （2） a communication overlapping method to overlap the communications between CPUs and GPUs, and （3） a dynamic workload balancing method to adjust the workloads among compute nodes, the implementation achieves a speedup of 4.16x on one NVIDIA Tesla M2050 GPU compared to a 2.93 GHz six-core Intel Xeon X5670 CPU. In addition, it runs 2.41x faster on 256 compute nodes of Tianhe-lA （with two CPUs and one GPU inside a node） than on 256 GPU-excluded nodes.展开更多
基金This work is supported by the National Natural Science Foundation of China under the Grants 62020106003,61922042,and 61773201the Fundamental Research Funds for the Central Universities of China under the Grants NC2020002 and NP2020103+2 种基金the 111 Project of the Programme of Introducing Talents of Discipline to Universities of China under the Grant B20007,the 13th Five-Year Equipment Pre-research Projects of China under the Grant 30501050403the Science Foundation of Science and Technology on Space Intelligent Control Laboratory of China under the Grant 6142208180201the Joint Foundation of Equipment Advance Research and Aerospace Science and Technology of China under the Grant 6141B060910.
文摘In this paper,the adaptive fault-tolerant formation tracking control problem for a set of heterogeneous unmanned aerial vehicle(UAV)and unmanned ground vehicle(UGV)systems with actuator loss of effectiveness faults is investigated.The cooperative fault-tolerant formation control strategy for UAV and UGV collaborative systems is classi-ed into the altitude consensus control scheme for follower UAVs and the position cooperative formation control scheme for all followers.The altitude consensus control algorithm is designed by utilizing backstepping control technique to drive all UAVs to a desired prede-ned height.Then,based on synchronization formation error information,the position cooperative formation control algorithm is proposed for all followers to reach the expected position and perform the desired formation con-guration.The adaptive fault estimation term is adopted in the designed fault-tolerant formation control algorithm to compensate for the actuator loss of effectiveness fault.Finally,a simulation example is proposed to reveal the validity of the designed cooperative formation tracking control scheme.
基金supported by the Science Foundation of LCPNational Natural Science Foundation of China(Grant Nos.11475028 and 11325209)+4 种基金the Opening Project of State Key Laboratory of Explosion Science and Technology(Beijing Institute of TechnologyGrant No.KFJJ14-1M)the Open Project Program of State Key Laboratory of Theoretical Physics,Institute of Theoretical PhysicsChinese Academy of SciencesChina(Grant No.Y4KF151CJ1)
文摘In this mini-review we summarize the progress of modeling, simulation and analysis of shock responses of heterogeneous materials in our group in recent years. The basic methodology is as below. We first decompose the problem into different scales. Construct/Choose a model according to the scale and main mechanisms working at that scale. Perform numerical simulations using the relatively mature schemes. The physical information is transferred between neighboring scales in such a way: The statistical information of results in smaller scale contributes to establishing the constitutive equation in larger one. Except for the microscopic Molecular Dynamics(MD) model, both the mesoscopic and macroscopic models can be further classified into two categories,solidic and fluidic models, respectively. The basic ideas and key techniques of the MD, material point method and discrete Boltzmann method are briefly reviewed. Among various schemes used in analyzing the complex fields and structures, the morphological analysis and the home-built software, GISO, are briefly introduced. New observations are summarized for scales from the larger to the smaller.
基金supported by National Natural Science Foundation of China(grant No.52176095)Anhui Provincial Natural Science Foundation(grant No.2308085ME189)China Postdoctoral Science Foundation(grant No.2019M661688).
文摘Water vapor nucleation on particle's surface plays an important role in dust removal,cloud formation,and particle measurement.However,the selectivity of nucleation sites and the nucleation characteristic of water molecule on the particle's surface are still unclear,especially for the aggregated particles.In this paper,the effects of particle wettability and aggregation modes on the selectivity of nucleation sites and the nucleation characteristics were investigated using molecular dynamics simulation.The results were compared with our earlier experimental findings.It illustrates how the contact angle of clusters,the growth velocity,and the growth duration are all influenced by the interaction coefficient between water and particles.Moreover,the nucleation sites of water molecules on the particle aggregation surface exhibit a definite selectivity.The primary indicator of this selectivity is the preferential nucleation of water molecules at the interfaces of linear chain aggregation particles,at the inner side of non-linear chain aggregation particles,and at the centers of ring aggregation.These results are in good agreement with our previous experimental findings.More significantly,additional research has revealed that subcritical-size clusters typically aggregate on two-particle surfaces spacing when the spacing smaller than the critical cluster size.
基金financially supported by the National Science Fund for Distinguished Young Scholars(No.20825416)the National Natural Science Foundation of China(No.21374051)+2 种基金973 program(No.2012CB821503)PCSIRT(No.IRT1257)CERS-1-61.A.C.S.acknowledges the support of the Natural Science and Engineering Council(NSERC)of Canada
文摘A two-stage transition upon crossing the glass transition of polystyrene with increasing temperature was precisely determined and interpreted by using solid-state nuclear magnetic resonance （SSNMR）, 1H-XH dipolar couplings based double quantum-filtered （DQF） and dipolar filter （DF） experiments and 13C chemical shift anisotropy （CSA） based centerband-only detection of exchange （CODEX） experiment are used to fully characterize the time scale of molecular motions during the glass transition. While differential scanning calorimetry （DSC） and CODEX experiment predicted the first stage of glass transiton, DQF and DF experiments provided the evidence for the second stage transition during which the time scale of molecular motions changed from very slow （t 〉 ms） to very fast （t 〈 Its）. The first stage of glass transition begins with the occurrence of remarkable slow re-orientation motions of the polymer backbone segments and ends when the degree of slow motion reaches maximum. The onset and endpoint of the conventional calorimetric glass transition of polystyrene can be quantitatively determined at the molecular level by SSNMR. In the second stage, a subsequent dramatic transition associated with the melting of the glassy components was observed. In this stage liquid-like NMR signals appeared and rapidly increased in intensity after a characteristic temperature Tf （-1.1Tg）. The signals associated with the glassy components completely disappeared at another characteristic temperature Tc （-1.2Tg）.
基金supported by the National Natural Science Foundation of China(21133006,21273104,91221301 and 91021010)the Ministry of Science and Technology(2013CB834601)
文摘Much progress has been achieved for both experimental and theoretical studies on the dissociative chemisorption of molecules on surfaces.Quantum state-resolved experimental data has provided unprecedented details for these fundamental steps in heterogeneous catalysis,while the quantitative dynamics is still not fully understood in theory.An in-depth understanding of experimental observations relies on accurate dynamical calculations,in which the potential energy surface and adequate quantum mechanical implementation are desired.This article summarizes the current methodologies on the construction of potential energy surfaces and the quantum mechanical treatments,some of which are promising for future applications.The challenges in this field are also addressed.
基金supported by the National Natural Science Foundation of China(Grant No.11371049)Fundamental Research Funds for the Central Universities(Grant No.2016JBM070)
文摘In opinion dynamics,the convergence of the heterogeneous Hegselmann-Krause(HK) dynamics has always been an open problem for years which looks forward to any essential progress.In this short note,we prove a partial convergence conclusion of the general heterogeneous HK dynamics.That is,there must be some agents who will reach static states in finite time,while the other opinions have to evolve between them with a minimum distance if all the opinions does not reach consensus.And this result leads to the convergence of several special cases of heterogeneous HK dynamics,including when the minimum confidence bound is large enough,the initial opinion difference is small enough,and so on.
文摘The results of molecular-dynamics (MD) simulation are obtained for structural and thermodynamic properties of the molten system, Na1-xPbx (x ≤ 0.1), at 698 K in the model of nearly free electronic gas (NFE approximation). The all numerical experiments are carried out by unified procedure: 1) equilibrating the MD cell 1 ps;2) calculating partial radial distribution functions gab(r) in 1, 5, and 10 ps;3) data handling for calculating other characteristics. It follows from this that lead impurity in liquid sodium at concentration in the range of 1% - 10% is characterized by micro-heterogenetic structure in the form of cluster compounds with variable composition. Just therefore the sodium alloys, Na-MIV, with four-group elements of Periodic table have no eutectic in this range of additive concentrations. This eutectic is needed for modifying sodium coolant of the fast nuclear reactor. Therefore it is reasonable to find an alternative alloy of sodium with additive from adjacent groups which has a eutectic in this range of concentrations, for example, the eutectic, Na0.929Tl0.071, with melting point of 64°C. The modified sodium coolant by isotope, 205Tl, can appear attractive for inhibiting the chemical activity of sodium just as the lead alloyed one.
基金Supported by the National Natural Science Foundation of China (Nos. 61170049 and 60903044)the National High-Tech Research and Development (863) Program of China (Nos. 2012AA01A301 and 2012AA010903)
文摘Heterogeneous systems with both Central Processing Units （CPUs） and Graphics Processing Units （GPUs） are frequently used to accelerate short-ranged Molecular Dynamics （MD） simulations. The most time-consuming task in short-ranged MD simulations is the computation of particle-to-particle interac- tions. Beyond a certain distance, these interactions decrease to zero. To minimize the operations to investi- gate distance, previous works have tiled interactions by employing the spatial attribute, which increases the memory access and GPU computations, hence decreasing performance. Other studies ignore the spatial attribute and construct an all-versus-all interaction matrix, which has poor scalability. This paper presents an improved algorithm. The algorithm first bins particles into voxels according to the spatial attributes, and then tiles the all-versus-all matrix into voxel-versus-voxel sub-matrixes. Only the sub-matrixes between neighbor- ing voxels are computed on the GPU. Therefore, the algorithm reduces the distance examine operations and limits additional memory access and GPU computations. This paper also adopts a multi-level program- ming model to implement the algorithm on multi-nodes of Tianhe-lA. By employing （1） a patch design to ex- ploit parallelism across the simulation domain, （2） a communication overlapping method to overlap the communications between CPUs and GPUs, and （3） a dynamic workload balancing method to adjust the workloads among compute nodes, the implementation achieves a speedup of 4.16x on one NVIDIA Tesla M2050 GPU compared to a 2.93 GHz six-core Intel Xeon X5670 CPU. In addition, it runs 2.41x faster on 256 compute nodes of Tianhe-lA （with two CPUs and one GPU inside a node） than on 256 GPU-excluded nodes.
