| 高翔宇，吴超，吴宗秀，冯钊，段萧.基于泊松卡尔曼滤波与质点滤波的水下航行器寻源方法[J].海洋工程,2026,(1):185~193 |
| 基于泊松卡尔曼滤波与质点滤波的水下航行器寻源方法 |
| Source-seeking method for underwater vehicles based on Poisson Kalman filter and point mass filter |
| 投稿时间：2024-10-23 修订日期：2024-11-24 |
| DOI：10.16483/j.issn.1005-9865.2026.01.017 |
| 中文关键词: 质点滤波 泊松卡尔曼滤波 贝叶斯状态估计 水下航行器 场源搜索 |
| 英文关键词:point mass filter Poisson Kalman filter Bayesian state estimation underwater vehicle source seeking |
| 基金项目:国家重点研发计划课题资助项目（2023YFC2809604）；国家自然科学基金国家重大科研仪器研制项目（部门推荐）（42327901） |
|
| 摘要点击次数: 20 |
| 全文下载次数: 0 |
| 中文摘要: |
| 以海洋物理化学场寻源为背景，针对水下航行器的场源搜索问题，提出了一种基于泊松卡尔曼滤波与质点滤波的场源参数估计方法。在贝叶斯状态估计的递归过程中，通过泊松卡尔曼滤波估计粒子释放速率参数，通过质点滤波更新场源坐标概率密度，实现场源参数估计的分离，减少了场源状态空间的维度，最后根据信息熵指导航行器的采样行为。分别在二维和三维空间寻源场景中进行了仿真验证，结果显示随着采样次数的增加，该方法对粒子释放速率参数的估计值可以较好地收敛到真实值附近，且可以保证场源位置概率密度估计的准确性。与传统粒子滤波算法仿真结果对比表明，所提方法在保证场源参数估计准确性的同时可以显著减少粒子数量，达到降低维度、提高计算效率的目的。 |
| 英文摘要: |
| In the context of marine physical and chemical field source-seeking, a field source parameter estimation method based on the Poisson Kalman filter and the point mass filter is proposed for underwater vehicle source-seeking tasks. In the recursive process of Bayesian state estimation, the Poisson Kalman filter estimates the particle release rate parameter, while the point mass filter updates the probability density of the source's coordinates. This method achieves separation of field source parameter estimation and reduces the dimensionality of the field source state space . Finally, information entropy is applied to guide the vehicle's sampling behavior. Simulation validations were conducted in both two-dimensional and three-dimensional source-seeking scenarios, and results show that with an increasing number of samples, the estimated particle release rate parameter converges well towards the true value. Moreover, the method ensures accurate estimation of the probability density of the source location. Comparison with the traditional particle filter algorithm demonstrates that the proposed method significantly reduces the number of required particles, achieves dimensionality reduction, and improves computational efficiency while maintaining the accuracy of source parameter estimation. |
| 查看全文 查看/发表评论 下载PDF阅读器 |
| 关闭 |
