论文部分内容阅读
在非线性降维算法Isomap的基础上进行了改进,提出了一种基于度量多维标定法的空间变换方法。将原始网络空间中的路网距离转换为新欧氏空间中的近似路网距离,并在此距离度量基础上实现Kriging方法。通过对南昌市真实数据进行交通状态估计的实验发现,该方法比现有的基于欧氏距离度量的Kriging方法具有更高的估计精度,能够有效地解决交通领域中大规模路网交通运行状态监控的问题。
Based on the non-linear dimensionality reduction algorithm Isomap, an improved spatial transformation method based on the metric multi-dimension calibration method is proposed. Converting the road network distance in the original cyberspace to the approximate road network distance in the new Euclidean space and implementing the Kriging method based on this distance measure. Through the experiment of traffic status estimation of real data in Nanchang City, it is found that this method has higher estimation accuracy than the existing Kriging method based on the Euclidean distance measure, which can effectively solve the problems of large-scale traffic network monitoring The problem.