一种高速激波风洞实验数据与CFD仿真智能同化方法
By constructing a deep reinforcement learning agent to interact with the CFD simulation environment and automatically adjusting the CFD model parameters, the problem of sparse and noisy experimental data for high-speed aircraft is solved, achieving efficient and accurate CFD model assimilation, which is applicable to different wind tunnel equipment and aircraft.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- NORTHWESTERN POLYTECHNICAL UNIV
- Filing Date
- 2026-05-07
- Publication Date
- 2026-07-17
AI Technical Summary
In existing high-speed aircraft research, experimental time is extremely short, measurement data is sparse and noisy, sensor data is difficult to obtain, and CFD model parameters are difficult to determine. There is a need for a method that can automatically and efficiently solve this problem.
We construct a deep reinforcement learning agent, extract steady-state features through a temporal adversarial network, build an interactive CFD simulation environment, define a reward function, enable the agent to train interactively with the CFD environment, automatically adjust the CFD model parameters, and achieve data assimilation.
It achieves efficient utilization of sparse experimental data, automatically adjusts CFD model parameters, reduces the computational load of traditional methods, provides uncertainty quantification, is applicable to different wind tunnel equipment and aircraft shapes, and improves the accuracy of simulation results.
Smart Images
Figure CN122133530B_ABST