当前位置: X-MOL 学术Int. J. Pattern Recognit. Artif. Intell. › 论文详情
Our official English website, www.x-mol.net, welcomes your feedback! (Note: you will need to create a separate account there.)
Parameter Optimization of Impedance Gradient Change Medium Based on Reinforcement Learning
International Journal of Pattern Recognition and Artificial Intelligence ( IF 0.9 ) Pub Date : 2020-08-31 , DOI: 10.1142/s0218001421510022
Ke Li 1 , Shengjun Li 2 , Zhonghua Bao 3 , Qianqian Liu 3 , Tao Liu 3
Affiliation  

Based on the relative researches, in order to solve the problem that the parameters of impedance gradient change medium are difficult to be optimized and generalized in different environments, an optimization method of the parameters of the impedance gradient change medium based on reinforcement learning was proposed. First, the propagation principle of sound wave in impedance gradient medium was analyzed. The sound field distribution in the medium was also studied, in order to master its acoustic characteristics. Second, the parameters of sound velocity and impedance distribution were optimized by DQN algorithm to reduce the sound reflection. Finally, the effectiveness of the proposed reinforcement learning model was verified by the traditional method. The experimental results show that the method presented in this paper was superior to the traditional method. The trained parameters are effective to reduce the acoustic reflection to a lower level.

中文翻译:

基于强化学习的阻抗梯度变化介质参数优化

在相关研究的基础上,为解决阻抗梯度变化介质参数在不同环境下难以优化和泛化的问题,提出了一种基于强化学习的阻抗梯度变化介质参数优化方法。首先,分析了声波在阻抗梯度介质中的传播原理。还研究了介质中的声场分布,以掌握其声学特性。其次,通过DQN算法优化声速和阻抗分布参数,以减少声音反射。最后,通过传统方法验证了所提出的强化学习模型的有效性。实验结果表明,本文提出的方法优于传统方法。训练后的参数可以有效地将声反射降低到较低的水平。
更新日期:2020-08-31
down
wechat
bug