论文部分内容阅读
这组系列文章将介绍作者研究提出的分析预测理论 ,以及该理论在目标空间位置探测数据处理问题中的应用。从宏观上讲这种应用几乎平行于传统方法地给出了一整套全新的空间位置探测数据理论和算法。从微观上看 ,这种应用还具有一些传统方法没有的特点 ,例如可以对目标运动的模式进行识别 (因此确保我们以与目标运动模式相同的或类似的模拟方程对目标进行滤波 ,从而获得精度更高的预测值 ) ,可以给出目标在未来时刻的真实值的范围为我们进行数据关联提供更好的判据 ,可以解决一维和两维探测数据的融合问题 ,可以对不同探测精度的数据进行点迹融合等。另外该系列文章研究的工作与传统的工作有较大的互补性。该系列文章首先给出空间位置探测数据的处理问题的定义 ,并从中抽象出关键技术。在接下来的系列文章中将讨论解决这些关键技术的思路以及关键技术的分解 ,理论数学模型 ,分析预测理论 ,分析预测处理方法的解的存在性证明 ,相关算法等。其核心有 ,目标运动模式的识别技术 (模拟方程的选择问题 ) ,模拟方程参数的选择技术 ,运动模式发生改变的识别技术 ,航迹聚类技术 ,航迹和点迹聚类技术 ,航迹融合技术 ,不同探测精度和维数的点迹融合技术 ,位置预测的评估技术等。最后给出仿真实验数据和结果 ,与传统方?
This series of articles will introduce the analysis and prediction theory proposed by the author and the application of the theory in the data processing of the target spatial location detection. From a macroscopic point of view this kind of application is almost parallel to the traditional method gives a whole new set of space position detection data theory and algorithm. From the microscopic point of view, this application also has some traits that traditional methods do not have, such as recognizing patterns of target motion (thus ensuring that we filter the target with the same or similar simulation equations as the target motion pattern to achieve accuracy Higher predictive value), the range of the true value of the target in the future can be given to provide better criteria for our data association, which can solve the problem of the fusion of one-dimensional and two-dimensional probing data, and can analyze the data of different probing accuracy For spot fusion and so on. In addition, the work of this series of articles has great complementarity with the traditional work. This series of articles first gives the definition of the processing of spatial location detection data, and abstract the key technologies. In the following series of articles, the ideas to solve these key technologies and the decomposition of key technologies, theoretical mathematical models, analysis and forecasting theories, proof of existence of solutions for predictive processing methods and related algorithms will be discussed. Its core is the recognition technology of target sports mode (selection of simulation equation), the selection technology of simulation equation parameters, the recognition technology of change of movement mode, track clustering technology, track and spot trace clustering technology, track Fusion technique, spot fusion technology with different detection accuracy and dimension, evaluation technology of position prediction and so on. Finally, the simulation experimental data and results, and the traditional side?