自主式交通系统框架功能域划分

Function domain classification of autonomous transportation system framework

可自主运行、演化的新一代自主式交通系统(ATS)发展亟需完善体系框架,以应对技术和需求的迅猛发展。以ATS框架中的功能为对象,旨在为系统功能建立一套客观且可验证推广的科学划分方法,使得ATS体系框架中的信息流动更加清晰,为系统仿真验证提供支持。通过定义7个维度的属性将功能描述转化为分类文本数据,改进K-modes聚类算法建立多属性分类数据聚类模型。聚类中心采用多个属性值及其类内出现的概率组合,并使用信息熵计算类内属性权重。最终,以相异度之和最小作为目标函数输出结果,并从迭代方式、属性选取等角度展开灵敏度分析。研究结果表明:所提方法基本满足功能域划分需求,通过有监督迭代、功能属性细化及描述规范化处理等手段进一步提升功能划分合理性。

The development of a new generation of autonomous transport systems(ATS)that can operate and evolve autonomously requires an improved system framework to cope with the rapid advancements in technology and demand.This article aims to establish an objective and verifiable scientific classification method for system functions within the ATS framework.The proposed method is intended to facilitate clearer information flow and support system simulation.To achieve this goal,functional descriptions are transformed into classified text data by defining seven dimensions of attributes.The multi-attribute classification data clustering model is then established by improving the K-modes clustering algorithm.The clustering center uses a combination of multiple attribute values and their intra-class occurrence probabilities.Additionally,information entropy is used to calculate the intra-class attribute weights.Finally,the objective function output is obtained by minimizing the sum of dissimilarity.Sensitivity analysis is carried out from the perspectives of the iteration method and attribute selection.The results demonstrate that the proposed method basically meets the requirements of functional domain classification.However,supervised iteration,refinement of functional attributes,and specification of functional role description can further improve the rationality of functional classification.