近岸水域悬浮物浓度的多尺度动态检测理论与关键技术研究

Underwater acoustic detection is one of the most important and fundamental technologies for ocean monitoring and development. With the continuous advancement of national marine power construction, marine pollution control and ecological environment restoration have become more and more urgent. Thus the demand of marine ecological and environmental monitoring technologies and equipments is increasing greatly. In regard of the situation with coastal waters serious pollution of suspended matter and urgent need of marine environmental monitoring technologies, concerning the transmission properties of sound waves, this project studies the detection model based on the multi-band frequency modulation circular interleaved array, parameter estimation method of approaching theoretical limits under strong color noise, multi-scale inversion technique of suspended matter concentration based on broadband acoustic comprehensive modeling of forward / backward scattering, development of GIS based visualization system using geographic information, so as to realize the theory and key technologies for visual online detection of dynamic multi-scale resolution of underwater suspended matter concentration. The research results are expected to solve the online monitoring problem of multi-scale resolution for suspended matter concentration in water, and improve the completeness and detection accuracy of underwater suspension detection data. Furthermore, they could provide abundant basic data for marine resources exploitation and environmental protection, develop marine detection technologies with independent intellectual property rights, as well as promote the prevention of marine pollution and development of marine industry.

水声检测技术是海洋监测与开发最重要和基础的关键技术之一。随着国家建设海洋强国的进程不断推进，海洋污染防治和生态环境修复日益紧迫，对海洋生态与环境监测技术和设备的需求也越发迫切。本项目针对近岸水域悬浮物污染严重和海洋环境监测技术亟需的现状，根据声波的传输特性，通过研究基于环状多频段调频超声交织阵列的建模、强色噪声下逼近理论极限界的幅度/功率谱及多普勒参数估计方法、基于宽带声学前/后向散射综合建模的悬浮物浓度多尺度反演算法和利用地理信息的可视化呈现，实现水下悬浮物浓度的多尺度动态可视化检测理论与关键技术。研究成果有望能够有效地解决水下不同粒径悬浮物浓度的在线监测问题，提高水下悬浮物检测的精准性和数据完备性，为海洋资源开发与海洋环境保护提供丰富的基础海洋数据，形成具有自主知识产权的海洋检测技术，推动海洋环境污染的防治以及海洋产业的发展。

本项目针对近海及河口水域内悬浮物污染严重和海洋环境监测技术亟需现状，根据声波传输特性，从理论上分析出基于调频阵列检测模型的参数估计理论界，实现了基于多频段调频阵的主动式水声检测装置、强色噪声环境下回波信号参数估计算法、多普勒估计算法、基于宽带声学前/后向综合建模的多尺度悬浮物三维流速和浓度反演算法、结合地理信息系统的可视化处理以及相关成果在水声检测等领域中的创新应用。项目创新性采用基于多高斯分布假设的贝叶斯估计和深度学习等方法实现了高准确率的悬浮物浓度的检测方法、利用阵列特性实现了与声速无关的水下目标定位方法等。本项目成果已经在国内外期刊与会议上发表科研论文12篇，其中SCI收录期刊论文6篇、EI收录期刊论文3篇、EI收录IEEE的国际会议上发表论文3篇、提交国际发明专利（PCT）申请2件、提交专利申请16件，其中国家发明专利申请15件，已经授权国家发明专利5件，实用新型专利授权1件、获得软件著作权1件、培养了研究生10名，其中6名研究生已获得学位、骨干研究人员1名获得职称提升。项目研究成果预期在水质检测、海洋环境监测等领域有良好的应用前景。