基于多源信息融合的ADCP流速全剖面精确重构

Acoustic Doppler Current Profiler (ADCP) has advantages of high precision and can obtain three-dimensional velocity profile so that it is widely adopted in flow velocity measurement and hydrological survey. However, due to complex measurement circumstances and influence on performances of instrument, in most cases it is difficult to acquire in a high precision and a full profile of measuring result. To this end, this project plans to reconstruct velocity profile on the basis of multi-source information fusion to solve this problem. The project firstly studies on systemic substitute methods of external boat speed and boat orientation reference, and according to that, a universal and accurate flow measurement method of ADCP based on information fusion of multisource external sensors has proposed. After that, the project studies on attitude change and deviation of sound velocity’s correction model of depth cell of velocity and spatial position of ADCP, according to that the flow velocity profile reconstruction method is put forward. What’s more, combining vertical gradient change rules of flow velocity with common velocity distribution models, a self-adaption determination method of flow velocity model in blind zone by different tense and flow regime is give out. Finally, combining reconstructed ADCP flow profile with flow velocity of blind zone that determined by self-adoption method, the full and high-precision profile of flow velocity is obtained. The project study improves the precision of ADCP measurement, expands the application of ADCP, and provides theoretical and methodological support for applications in hydraulic engineering and research of hydrologic sciences.

声学多普勒流速剖面仪（ADCP）具有测量速度快、可获得三维流速剖面等优势，在流速测量和水文测验中被广泛采用。但受复杂测量环境及仪器性能影响，ADCP测量结果往往难以全剖面、高精度获取。为此，项目在多源信息融合的基础上提出一种ADCP流速全剖面高精度重构方法。首先，替代现有ADCP流速测量的船速和方位基准数学模型，提出一种基于自适应卡尔曼滤波的多源信息融合方法，获取准确、稳健的船速和方位基准；其次，研究声速、姿态变化以及流场差异对实测流速剖面及其空间位置的影响机理，构建流速剖面精确重构模型；再次，根据重构的实测流速剖面的垂向梯度变化规律，研究不同时态、流态下上、下盲区流速的最优模型；最后，将重构的实测区流速剖面和最优模型确定的盲区流速融合，实现ADCP流速全剖面、高精度重构。项目研究成果提高ADCP的测量精度，拓展ADCP的应用范围，为水文科学研究及水利工程应用提供理论和方法支撑。

受仪器性能及复杂测量环境影响，声学多普勒流速剖面仪（ADCP）在实际应用中仍存在流速测量结果存在偏差及流速剖面不完整等问题。本项目通过基于多源信息融合的ADCP船速和方位基准精确确定、顾及声速/姿态变化及流场差异的ADCP实测区流速剖面精确重构以及基于流速垂向梯度变化的ADCP盲区流速精确估计等3个方面的研究，实现了ADCP流速的全剖面、高精度重构。主要研究成果包括：.（1）针对底质流动引起的ADCP船速不准及内置磁罗经精度偏低且易受外部磁场干扰带来的方位不准问题，给出了基于多源信息融合的ADCP流速精确确定的系统性解决方案和数据处理方法。实验结果表明，该方法完全消除了上述因素对流速测量的影响，并获得了±0.03 m/s及±3.8°的流速和流向测量精度，远优于传统方法的±0.06 m/s及±20.0°的测量精度。.（2）借助姿态和方位信息将波束坐标系下的流速值转换为地理坐标系下的绝对流速剖面后，然后对剖面中各水深单元位置插值重构，实现流速剖面的准确归位；在此基础上，基于相邻历元交会和多测次流速观测值，采用最小二乘方法实现了ADCP流速剖面的精确重构。.（3）总结了4种常用的流速分布模型，并针对这些模型待求参数较多、不能直接应用于ADCP盲区计算的不足，对各模型进行了简化并推导出了适用于盲区流速的推算模型；在此基础上，针对现有垂线流速变化具有单一性假设的缺陷，结合流速垂向梯度变化规律给出了一种ADCP盲区流速最优推算模型的自适应确定方法。实验表明，最优模型相较于传统的单一模型可以提高2%的流量估计精度。对于年均流量达到28700 m3/s的长江来说，这个改进仍然可观。.在本基金项目的资助下，项目负责人共发表了4篇期刊论文（其中2篇为SCI检索，1篇为EI检索）、1篇会议论文（SCI检索），撰写了1部专著，申请了1项国家发明专利，完成了1项软件著作权登记，完全达到了预期的研究计划。