基于存活概率的动态车龄分布模型

城市车辆的年龄组成(简称车龄分布)是体现城市车辆老化程度和确定车辆报废年限的重要指标。目前已有的车龄分布算法要求数据量大,在我国大部分城市很难得到完整的数据。笔者提出一种符合威布尔分布的车辆存活概率算法。该算法根据车辆保有量数据和报废车辆总数(或新车数据)推导出各年的车龄分布,且能够动态预测将来年份的车龄分布。同时,利用北京市小客数据对算法进行了实例应用,证明该法的实用可行性。由于该算法所需数据量小,计算简便,特别适用于我国在用车统计不完备的地区,利用该算法根据有限的数据来推导出当地的车龄分布,具有通用性。该算法的有效性检验和敏感度分析需进一步研究。

Survival Probability-based Dynamic Vehicle Age Distribution Model

Vehicle age distribution (VAD) is an important factor reflecting the vehicle aging degree and determining the scrapping time limit. The data needed by available VAD models are rather big which are difficult to obtain in most cities in China. A survival probability-based dynamic model following Weibull probability distribution is proposed for calculation and forecast of VAD, by which annual vehicle age distribution could be derived from total number of registered vehicles and total number of newly registered vehicles (or number of scrapped vehicles), and the age distribution in future years could be predicted dynamically. The model is exemplified by the data of cars in Beijing and is proved to be feasible showing the data needed is rather small and the calculation is rather simple. The model might be applied to those regions with limited data in where the vehicle registration information is incomplete or unavailable, and is generalized in some extent. The validity and sensitivity analysis need to be conducted in future.