【摘要】 无人机工作过程中会通过各类传感设备采集大量运行数据。数据中蕴含着无人机的重要运行参数,为了提升无人机参数分析的准确与可靠,针对无人机多源数据的时基、空基差异特性,提出了多源数据融合的参数智能感知方法。方法首先利用特征点、簇间距离与密度函数设计了多源数据度量集合,根据运行数据的似然估计以及高斯混合构造多源数据的融合方程及求解方式。然后设计了单源数据融合的质量评价方法,结合准确度、完整度与相似度建立融合数据的评价模型。最后针对无人机任务对性能参数的动态影响,对单源数据融合质量采取加权计算后,得到多源数据融合评估,确定最终的感知参数结果。仿真结果表明,多源数据融合的参数智能感知方法能够有效应对无人机运行数据的多源异构特性,提高多源数据融合质量和参数感知响应,且数据融合性能与数据源具有良好的动态稳定性。更多还原

【Abstract】 In order to improve the accuracy and reliability of UAV parameter analysis, a parameter intelligent sensing method of multi-source data fusion is proposed for the time-based and space-based difference characteristics of UAV multi-source data. Firstly, the measure set of multi-source data was designed by using feature points, the distance between clusters and density function. According to the likelihood estimation of running data and Gaussian mixture, the fusion equation and solution method of multi-source data were constructed. Then the quality evaluation method of single-source data fusion was designed, and the evaluation model of fusion data was established by combining accuracy, integrity and similarity. Finally, in view of the dynamic impact of the UAV task on performance parameters, after weighted calculation of single-source data fusion quality, the multi-source data fusion evaluation was obtained to determine the final perception parameter results. The simulation results show that the multi-source data fusion method can effectively deal with the heterogeneous characteristics of UAV operation data, improve the quality of multi-source data fusion and parameter sensing response, and the data fusion performance and data source have good dynamic stability.更多还原