Turbulence field update method and apparatus, and related device thereof
View Patent ↗Provided are a turbulence field update method, apparatus, and device, and a computer-readable storage medium. The method includes: obtaining sample turbulence data; performing model training by use of the sample turbulence data to obtain a reinforcement learning turbulence model; calculating initial turbulence data of a turbulence field by use of a Reynolds Averaged Navior-Stokes (RANS) equation; processing the initial turbulence data by use of the reinforcement learning turbulence model to obtain a predicted Reynolds stress; and performing calculation on the predicted Reynolds stress by use of the RANS equation to obtain updated turbulence data.
1. A turbulence field update method, comprising:
obtaining sample turbulence data;
performing model training by use of the sample turbulence data to obtain a reinforcement learning turbulence model;
calculating initial turbulence data of the turbulence field by use of a Reynolds Averaged Navier-Stokes (RANS) equation;
processing the initial turbulence data by use of the reinforcement learning turbulence model to obtain a predicted Reynolds stress; and
using the predicted Reynolds stress as an input to the RANS equation to obtain updated turbulence data;
wherein the performing model training by use of the sample turbulence data to obtain a reinforcement learning turbulence model comprises:
performing feature extraction on the sample turbulence data to obtain a sample turbulence feature; and
performing model training by use of the sample turbulence feature to obtain the reinforcement learning turbulence model.
2. The turbulence field update method according to claim 1 , wherein the performing model training by use of the sample turbulence feature to obtain the reinforcement learning turbulence model comprises:
constructing a Deep-Q-Network (DQN); and
performing iterative training on the DON by use of the sample turbulence feature with a preset target function as an iteration condition to obtain the reinforcement learning turbulence model.
3. The turbulence field update method according to claim 1 , wherein the calculating initial turbulence data of the turbulence field by use of a RANS equation comprises:
obtaining an initial Reynolds stress of the turbulence field; and
closing a Reynolds stress item of the RANS equation by use of the initial Reynolds stress to calculate the initial turbulence data.
4. The turbulence field update method according to claim 1 , wherein the processing the initial turbulence data by use of the reinforcement learning turbulence model to obtain a predicted Reynolds stress comprises:
obtaining a learning experience and network parameter during training of the reinforcement learning turbulence model;
optimizing the reinforcement learning turbulence model by use of the learning experience and the network parameter to obtain an optimized reinforcement learning turbulence model; and
processing the initial turbulence data by use of the optimized reinforcement learning turbulence model to obtain the predicted Reynolds stress.
5. A turbulence field update device, comprising:
a memory, configured to store a computer program;
a processor, configured to execute the computer program to implement any steps of a turbulence field update method, wherein the turbulence field update method comprises:
obtaining sample turbulence data;
performing model training by use of the sample turbulence data to obtain a reinforcement learning turbulence model;
calculating initial turbulence data of the turbulence field by use of a Reynolds Averaged Navier-Stokes (RANS) equation;
processing the initial turbulence data by use of the reinforcement learning turbulence model to obtain a predicted Reynolds stress;
using the predicted Reynolds stress as an input to the RANS equation to obtain updated turbulence data;
wherein the performing model training by use of the sample turbulence data to obtain a reinforcement learning turbulence model comprises:
performing feature extraction on the sample turbulence data to obtain a sample turbulence feature; and
performing model training by use of the sample turbulence feature to obtain the reinforcement learning turbulence model.
6. The turbulence field update device according to claim 5 , wherein the performing model training by use of the sample turbulence feature to obtain the reinforcement learning turbulence model comprises:
constructing a Deep-Q-Network (DQN); and
performing iterative training on the DQN by use of the sample turbulence feature with a preset target function as an iteration condition to obtain the reinforcement learning turbulence model.
7. The turbulence field update device according to claim 5 , wherein the calculating initial turbulence data of the turbulence field by use of a RANS equation comprises:
obtaining an initial Reynolds stress of the turbulence field; and
closing a Reynolds stress item of the RANS equation by use of the initial Reynolds stress to calculate the initial turbulence data.
8. The turbulence field update device according to claim 5 , wherein the processing the initial turbulence data by use of the reinforcement learning turbulence model to obtain a predicted Reynolds stress comprises:
obtaining a learning experience and network parameter during training of the reinforcement learning turbulence model;
optimizing the reinforcement learning turbulence model by use of the learning experience and the network parameter to obtain an optimized reinforcement learning turbulence model; and
processing the initial turbulence data by use of the optimized reinforcement learning turbulence model to obtain the predicted Reynolds stress.
9. A non-transitory computer-readable storage medium, having a computer program stored thereon which, when executed by a processor, implements any steps of a turbulence field update method, wherein the turbulence field update method comprises:
obtaining sample turbulence data;
performing model training by use of the sample turbulence data to obtain a reinforcement learning turbulence model;
calculating initial turbulence data of the turbulence field by use of a Reynolds Averaged Navier-Stokes (RANS) equation;
processing the initial turbulence data by use of the reinforcement learning turbulence model to obtain a predicted Reynolds stress;
using the predicted Reynolds stress as an input to the RANS equation to obtain updated turbulence data;
wherein the performing model training by use of the sample turbulence data to obtain a reinforcement learning turbulence model comprises:
performing feature extraction on the sample turbulence data to obtain a sample turbulence feature; and
performing model training by use of the sample turbulence feature to obtain the reinforcement learning turbulence model.
10. The non-transitory computer-readable storage medium according to claim 9 , wherein the performing model training by use of the sample turbulence feature to obtain the reinforcement learning turbulence model comprises:
constructing a Deep-Q-Network (DQN); and
performing iterative training on the DON by use of the sample turbulence feature with a preset target function as an iteration condition to obtain the reinforcement learning turbulence model.
11. The non-transitory computer-readable storage medium according to claim 9 , wherein the calculating initial turbulence data of the turbulence field by use of a RANS equation comprises:
obtaining an initial Reynolds stress of the turbulence field; and
closing a Reynolds stress item of the RANS equation by use of the initial Reynolds stress to calculate the initial turbulence data.
12. The turbulence field update method according to claim 1 , wherein before the performing model training by use of the sample turbulence data to obtain a reinforcement learning turbulence model, the turbulence field update method further comprises:
preprocessing the sample turbulence data to obtain standard sample turbulence data.
13. The turbulence field update method according to claim 12 , wherein the preprocessing the sample turbulence data to obtain standard sample turbulence data comprises:
normalizing the sample turbulence data to obtain normalized sample turbulence data; and
extracting the normalized sample turbulence data at equal intervals to obtain the standard sample turbulence data.
14. The turbulence field update device according to claim 5 , wherein before the performing model training by use of the sample turbulence data to obtain a reinforcement learning turbulence model, the turbulence field update method further comprises:
preprocessing the sample turbulence data to obtain standard sample turbulence data.
15. The turbulence field update device according to claim 14 , wherein the preprocessing the sample turbulence data to obtain standard sample turbulence data comprises:
normalizing the sample turbulence data to obtain normalized sample turbulence data; and
extracting the normalized sample turbulence data at equal intervals to obtain the standard sample turbulence data.
16. The non-transitory computer-readable storage medium according to claim 9 , wherein before the performing model training by use of the sample turbulence data to obtain a reinforcement learning turbulence model, the turbulence field update method further comprises:
preprocessing the sample turbulence data to obtain standard sample turbulence data.
17. The non-transitory computer-readable storage medium according to claim 16 , wherein the preprocessing the sample turbulence data to obtain standard sample turbulence data comprises:
normalizing the sample turbulence data to obtain normalized sample turbulence data; and
extracting the normalized sample turbulence data at equal intervals to obtain the standard sample turbulence data.