检索
E-mail
RSS

作者投稿 主编审稿 专家审稿 远程编辑

热带海洋学报

• • 上一篇    

基于RLPSO算法的卷积神经网络地震P波识别方法研究

王罗文1, 2, 3, 杨泽荣4, 文永鹏4, 朱心科2, 3, 颜波1, 2, 3, 秦华伟1   

  1. 1. 杭州电子科技大学机械工程学院, 浙江 杭州 310018;

    2. 海底科学与划界全国重点实验室, 浙江 杭州 310012;

    3. 自然资源部第二海洋研究所, 浙江 杭州 310012;

    3. 国电电力浙江舟山海上风电开发有限公司, 浙江 舟山 316000

  • 收稿日期:2025-03-11 修回日期:2025-04-11 接受日期:2025-04-18
  • 通讯作者: 秦华伟
  • 基金资助:

    国家重点研发计划 (2022YFC3003802); 自然资源部海洋观测技术重点实验室研究基金(2024klootA11); 自然资源部海底科学重点实验室开放基金(KLSG2409)

Research on Seismic P - wave Identification Method Based on Convolutional Neural Network Optimized by RLPSO Algorithm

WANG Luowen1, 2, 3, YANG Zerong4, WEN Yongpeng4, ZHU Xinke2, 3, YAN Bo1, 2, 3, QIN Huawei1   

  1. 1. School of Mechanical Engineering, Hangzhou Dianzi University, Hangzhou 310018, China;

    2. State Key Laboratory of Submarine Geoscience, Hangzhou 310012, China;

    3. Second Institute of Oceanography, Hangzhou 310012, China;

    4. Guodian Electric Power Zhejiang Zhoushan Offshore WindPower Development Co.,Ltd, Zhoushan 316000, China

  • Received:2025-03-11 Revised:2025-04-11 Accepted:2025-04-18
  • Supported by:

     National Key R&D Program of China(2022YFC3003802); the Research Fund of the Key Laboratory of Marine Observation Technology, Ministry of Natural Resources(2024klootA11); the Open Fund of the Key Laboratory of Submarine Geoscience, Ministry of Natural Resources(KLSG2409)

PDF (PC)

赞

摘要： 针对海洋地震P波实时监测系统的地震P波识别需求，设计了一种基于深度学习的地震P波识别算法。该算法通过构建卷积神经网络和强化学习粒子群优化算法的混合地震P波识别模型，利用先验地震数据对模型进行训练，并采用RLPSO (reinforcement learning particle swarm optimization)算法对卷积神经网络模型中的超参数进行迭代优化。将优化后的超参数代入CNN (convolutional neural network)模型后，实现了地震P波的识别。为了验证算法的有效性，采用测试数据集对不同算法进行了仿真测试。实验结果表明，该算法在训练过程中表现出更低的训练损失和更高的识别准确率。此外，该算法在低信噪比条件下仍能保持较高的识别精度，展现了较强的鲁棒性，满足了海洋地震P波实时监测系统的地震P波识别需求。

关键词: 地震P波识别, 卷积神经网络, 强化学习, 粒子群优化算法

Abstract: In response to the demand for real-time monitoring of seismic P-waves in marine environments, a deep learning-based algorithm for seismic P-wave identification has been designed. This algorithm constructs a hybrid model for seismic P-wave identification by integrating a CNN(convolutional neural network) with a RLPSO(reinforcement learning particle swarm optimization) algorithm. The model is trained using a prior seismic data as prior information, and the RLPSO algorithm is employed to iteratively optimize the hyperparameters of the CNN. The optimized hyperparameters are then incorporated into the CNN model to achieve predict the arrival time of seismic P-wave identifications. To validate the effectiveness of the algorithm, simulation tests were conducted on a test dataset using different algorithms. The experimental results demonstrate that the proposed algorithm exhibits lower training loss and higher identification accuracy during the training process. Furthermore, the algorithm maintains high identification precision under low signal-to-noise ratio conditions, demonstrating showcasing strong robustness. The algorithm meets the requirements for real-time seismic P-wave identification in marine seismic monitoring systems.

Key words: Seismic P-wave identification, Convolutional Neural Network, Reinforcement Learning, Particle Swarm Optimization Algorithm