论文部分内容阅读
路面结构设计中,要考虑设计年限内车辆对路面的综合累计损伤作用,必须对现有的交通量、交通组成、轴载组成以及增长规律进行分析和预估。目前获取这些数据的系统价格昂贵,且操作复杂,很难实现小规模、大范围的普及应用。FCM算法是一种以隶属度来确定每个数据点属于某个聚类程度的算法,由于能更准确描述模式间的不确定关系,摒弃了k-means等传统硬聚类算法初始点选择及聚类结果不稳定的缺点。基于大量高速公路实测轴载调查的数据,通过分析车辆的种类、轴型及汽车动静荷载对路面的影响,采用FCM聚类算法有效确定了整车重量下各级轴载及数学模型。数据聚类计算结果表明:经过该算法计算的数据可以很好地满足工程设计数据分析的需要,方便快捷,将为大范围内路面养护、维修以及新建路面的结构设计提供数据参考。
Pavement structure design, to consider the cumulative cumulative damage during the design of the vehicle on the road, the existing traffic volume, traffic composition, the composition of the axle load and the growth of the law to be analyzed and estimated. At present, the system for obtaining these data is expensive and complicated to operate, which makes it difficult to achieve small-scale and large-scale popularization. FCM algorithm is an algorithm to determine the degree of membership of each data point belongs to a certain degree of clustering. Because of the more accurate description of the relationship between the model uncertainty, and to abandon the k-means and other traditional hard clustering algorithm initial point selection and The disadvantage of clustering results is not stable. Based on the data of a large number of surveyed axles in expressways, the paper analyzes the influence of vehicle type, axle type and the static and dynamic load on the road surface, and uses FCM clustering algorithm to effectively determine the axle loading and mathematical models at all levels of vehicle weight. The results of data clustering show that the data calculated by this algorithm can meet the needs of engineering design data analysis quickly and conveniently, and provide data reference for pavement maintenance, maintenance and structural design of new pavement in a wide range.