基于船舶AIS及签证数据的长江内河货运阻抗及分布模型研究

摘要

为了研究长江内河各航段复杂多变的航道环境对水路货物运输的影响,根据船舶AIS数据及三峡船闸过闸数据建立一种考虑不同月份、不同航段的船舶逆水而上、顺流而下2种情况的货运阻抗函数,并基于Matlab求解得到长江内河重要港口节点之间不同月份的时间阻抗矩阵.以2010年10月船舶签证数据为例,结合距离阻抗矩阵和时间阻抗矩阵,标定阻抗函数为幂函数和指数函数下的重力模型的参数,建立相应的货运分布模型.并在不同模型下分别预测了2012年10月货运量分布情况.对货运实际观测值和模型预测值结果进行货运距离分布(trip length distribution,TLD)方法检验.结果显示,以时间为阻抗指标的幂函数和指数函数重力模型预测结果的重合率(coincidence ratios,CR)值分别为0.83和0.86,比对应的以距离为阻抗指标的预测结果的CR值0.81和0.82要高,所建立的以时间阻抗模型更能反映长江内河货运实际情况.%This study aims to figure out dynamic and complex impacts of various sections of Yangtze River on waterway freight transportation.Based on AIS and residence time of ships at the Three-Gorges Dam, a freight impedance model considering ship course, waterway segment and time is developed.Meanwhile, a time impedance matrix among important ports of the Yangtze River in different months is obtained using Matlab.Taking data of ship visa in October 2010 as a case study, the parameters of two types of gravity distribution models based on distance impedance matrix and time impedance matrix are calibrate respectively.Then corresponding models of distribution of freight transportation are established.With these models, the distribution in October 2012 is predicted.A Trip Length Distribution (TLD) method is applied to check the four predicted distributions with the actual distribution in October 2012.The results show that the Coincidence Ratios (CR) values got from the inverse power function and the exponential function based on the time impedance are 0.83 and 0.86 respectively;which based on the distance impedance are 0.81 and 0.82 respectively.This study indicates that the gravity distribution model based on time impedance has a better predication performance of inland freight transportation of the Yangtze River.