Jiaqi Liu
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Previous Organizations / Institutes
The organizations / institutes from the scholar's other articles in the past ten years
Research Subjects
Jiaqi Liu has a total of 21 co-authors. Here are the top co-authors by number of shared publications:
- Rongqian Chen
Xiamen University
3 shared publications Yancheng You
2 shared publications- Yancheng You
Xiamen University
2 shared publications Zhengwu Chen
2 shared publications- Danni Ai
Beijing Institute of Technology
1 shared publication - Deqiang Xiao
Beijing Institute of Technology
1 shared publication
Latest Publications
- 26 January 2025
- Published by Elsevier in Expert Systems with Applications
- Vol. 271, 126619
Incremental energy-based recurrent transformer-KAN for time series deformation simulation of soft tissue
- 26 January 2025
- Published by Elsevier in Expert Systems with Applications
- Vol. 271, 126619
Incremental energy-based recurrent transformer-KAN for time series deformation simulation of soft tissue
- 14 November 2023
- Published by Elsevier in Aerospace Science and Technology
- Vol. 143
Airfoils optimization based on deep reinforcement learning to improve the aerodynamic performance of rotors
- 14 November 2023
- Published by Elsevier in Aerospace Science and Technology
- Vol. 143
Airfoils optimization based on deep reinforcement learning to improve the aerodynamic performance of rotors
- 1 March 2023
- Published by AIP Publishing in Physics of Fluids
- Vol. 35 (3) , 037128
Aerodynamic optimization of airfoil based on deep reinforcement learning
The traditional optimization of airfoils relies on, and is limited by, the knowledge and experience of the designer. As a method of intelligent decision-making, reinforcement learning can be used for such optimization through self-directed learning. In this paper, we use the lift–drag ratio as the objective of optimization to propose a method for the aerodynamic optimization of airfoils based on a combination of deep learning and reinforcement learning. A deep neural network (DNN) is first constructed as a surrogate model to quickly predict the lift–drag ratio of the airfoil, and a double deep Q-network (double DQN) algorithm is then designed based on deep reinforcement learning to train the optimization policy. During the training phase, the agent uses geometric parameters of the airfoil to represent its state, adopts a stochastic policy to generate optimization experience, and uses a deterministic policy to modify the geometry of the airfoil. The DNN calculates changes in the lift–drag ratio of the airfoil as a reward, and the environment constantly feeds the states, actions, and rewards back to the agent, which dynamically updates the policy to retain positive optimization experience. The results of simulations show that the double DQN can learn the general policy for optimizing the airfoil to improve its lift–drag ratio to 71.46%. The optimization policy can be generalized to a variety of computational conditions. Therefore, the proposed method can rapidly predict the aerodynamic parameters of the airfoil and autonomously learn the optimization policy to render the entire process intelligent.
- 1 March 2023
- Published by AIP Publishing in Physics of Fluids
- Vol. 35 (3) , 037128
Aerodynamic optimization of airfoil based on deep reinforcement learning
The traditional optimization of airfoils relies on, and is limited by, the knowledge and experience of the designer. As a method of intelligent decision-making, reinforcement learning can be used for such optimization through self-directed learning. In this paper, we use the lift–drag ratio as the objective of optimization to propose a method for the aerodynamic optimization of airfoils based on a combination of deep learning and reinforcement learning. A deep neural network (DNN) is first constructed as a surrogate model to quickly predict the lift–drag ratio of the airfoil, and a double deep Q-network (double DQN) algorithm is then designed based on deep reinforcement learning to train the optimization policy. During the training phase, the agent uses geometric parameters of the airfoil to represent its state, adopts a stochastic policy to generate optimization experience, and uses a deterministic policy to modify the geometry of the airfoil. The DNN calculates changes in the lift–drag ratio of the airfoil as a reward, and the environment constantly feeds the states, actions, and rewards back to the agent, which dynamically updates the policy to retain positive optimization experience. The results of simulations show that the double DQN can learn the general policy for optimizing the airfoil to improve its lift–drag ratio to 71.46%. The optimization policy can be generalized to a variety of computational conditions. Therefore, the proposed method can rapidly predict the aerodynamic parameters of the airfoil and autonomously learn the optimization policy to render the entire process intelligent.
- Open Access
- 6 June 2022
- Published by Wiley in International Journal of Aerospace Engineering
- Vol. 2022, 1-19
Active Flow Control of Helicopter Rotor Based on Coflow Jet
When a helicopter rotor undergoes flow separation, the drag of the rotor increases substantially, as does the power demand, which seriously affects the aerodynamic performance and flight safety of the helicopter. Therefore, it is crucial to research how to suppress the flow separation of rotor blades. An active control technique based on a coflow jet (CFJ) at the rotor blade tip was employed in this study to suppress the flow separation over the rotor. The mechanisms and behavior were investigated. The rotor flow field was numerically simulated by solving the Reynolds-averaged Navier–Stokes equations with the finite volume method. The turbulence model was SST, and the rotor motion was simulated using the overset mesh technique. After applying CFJ control, the airflow from the injection slot at the leading edge of the rotor increased the energy of the mainstream in the near-wall area, which enhanced its ability to resist the adverse pressure gradient. A flow separation was effectively suppressed, both on the advancing and retreating sides, which improved the aerodynamic performance of the rotor. During the whole rotation period, the thrust coefficient of the rotor increased by up to 5.6%, the moment coefficient decreased by as much as 26.8%, and the equivalent lift-to-drag ratio increased by up to 44.0%. Moreover, the effects of the CFJ parameters on the flow separation suppression of the rotor are researched. These results may provide a foundation for the development of aerodynamic performance improvement for helicopter rotor based on a CFJ.
- 6 June 2022
- Published by Wiley in International Journal of Aerospace Engineering
- Vol. 2022, 1-19
Open AccessActive Flow Control of Helicopter Rotor Based on Coflow Jet
When a helicopter rotor undergoes flow separation, the drag of the rotor increases substantially, as does the power demand, which seriously affects the aerodynamic performance and flight safety of the helicopter. Therefore, it is crucial to research how to suppress the flow separation of rotor blades. An active control technique based on a coflow jet (CFJ) at the rotor blade tip was employed in this study to suppress the flow separation over the rotor. The mechanisms and behavior were investigated. The rotor flow field was numerically simulated by solving the Reynolds-averaged Navier–Stokes equations with the finite volume method. The turbulence model was SST, and the rotor motion was simulated using the overset mesh technique. After applying CFJ control, the airflow from the injection slot at the leading edge of the rotor increased the energy of the mainstream in the near-wall area, which enhanced its ability to resist the adverse pressure gradient. A flow separation was effectively suppressed, both on the advancing and retreating sides, which improved the aerodynamic performance of the rotor. During the whole rotation period, the thrust coefficient of the rotor increased by up to 5.6%, the moment coefficient decreased by as much as 26.8%, and the equivalent lift-to-drag ratio increased by up to 44.0%. Moreover, the effects of the CFJ parameters on the flow separation suppression of the rotor are researched. These results may provide a foundation for the development of aerodynamic performance improvement for helicopter rotor based on a CFJ.
- Open Access
- 15 December 2020
- Published by Wiley in International Journal of Aerospace Engineering
- Vol. 2020, 1-15
Study on Dynamic Stall Control of Rotor Airfoil Based on Coflow Jet
In this study, a dynamic stall control strategy, called the co-flow jet (CFJ), is applied to the rotor airfoil. The effect of the CFJ on the unsteady dynamic stall characteristics of the rotor airfoil is numerically investigated via numerical simulations of the unsteady Reynolds-averaged Navier-Stokes (URANS) equations coupled with the Spalart-Allmaras (S-A) turbulence model. The numerical methods are validated by a NACA0012 pitching airfoil case and a NACA6415 airfoil case based on the CFJ, and good agreement with experiments is found. Via the study of the typical conditions of CFJ control to suppress the dynamic stall of the OA212 rotor airfoil, it is verified that this method has a good effect on dynamic stall suppression. The diffusion and blending of the turbulent shear layer between the CFJ injection jet and the main flow excite the main flow and enhance its ability to resist the reverse pressure gradient; this suppresses the generation and development of the separation vortex, thereby enhancing the aerodynamic characteristics, improving the hysteresis effect, and increasing the system stability. On this basis, the control parameters of the CFJ are further studied, including the influences of the jet momentum coefficient and the positions and sizes of the injection and suction slots on suppressing the dynamic stall of the rotor airfoil. It is found that there is a jet momentum coefficient that optimizes the suppression effect of the dynamic stall of the rotor airfoil. Moreover, the position of the injection slot is found to have a greater effect on the dynamic stall suppression, while the size of the injection slot and the position and size of the suction slot have little effect.
- 15 December 2020
- Published by Wiley in International Journal of Aerospace Engineering
- Vol. 2020, 1-15
Open AccessStudy on Dynamic Stall Control of Rotor Airfoil Based on Coflow Jet
In this study, a dynamic stall control strategy, called the co-flow jet (CFJ), is applied to the rotor airfoil. The effect of the CFJ on the unsteady dynamic stall characteristics of the rotor airfoil is numerically investigated via numerical simulations of the unsteady Reynolds-averaged Navier-Stokes (URANS) equations coupled with the Spalart-Allmaras (S-A) turbulence model. The numerical methods are validated by a NACA0012 pitching airfoil case and a NACA6415 airfoil case based on the CFJ, and good agreement with experiments is found. Via the study of the typical conditions of CFJ control to suppress the dynamic stall of the OA212 rotor airfoil, it is verified that this method has a good effect on dynamic stall suppression. The diffusion and blending of the turbulent shear layer between the CFJ injection jet and the main flow excite the main flow and enhance its ability to resist the reverse pressure gradient; this suppresses the generation and development of the separation vortex, thereby enhancing the aerodynamic characteristics, improving the hysteresis effect, and increasing the system stability. On this basis, the control parameters of the CFJ are further studied, including the influences of the jet momentum coefficient and the positions and sizes of the injection and suction slots on suppressing the dynamic stall of the rotor airfoil. It is found that there is a jet momentum coefficient that optimizes the suppression effect of the dynamic stall of the rotor airfoil. Moreover, the position of the injection slot is found to have a greater effect on the dynamic stall suppression, while the size of the injection slot and the position and size of the suction slot have little effect.