摘要
汽车的智能驾驶决策问题是近年来机器学习领域的研究热点之一。驾驶员在驾驶过程中,通常会受到多种环境因素的影响,倘若驾驶员无法快速处理这些信息并做出正确的判断,无疑会引起安全事故。为此,在合理简化的前提下,基于驾驶决策行为以及影响驾驶决策相关环境因素的研究分析,将模糊理论和软间隔支持向量机理论相结合,提出了基于模糊支持向量机的单向变速决策模型。该模型旨在针对实时周边环境,为车辆驾驶员提供变速决策建议,帮助驾驶员更加高效地制定出安全的驾驶策略。仿真实验结果表明,相较于传统的支持向量机模型,建立于隶属函数之上的模糊支持向量机模型具有较高的决策正确率和较好的可拓展性,可为车辆驾驶员提供合理、安全的决策建议。
Automatic driving decision-making problem is one of the hot spots in the field of machine learning in recent years. The driv- ers,in the driving process,are usually affected by many environmental factors. And if a driver cannot quickly deal with the information and make the right decisions, he would undoubtedly suffer huge security risks. In order to solve this problem, on the analysis of the driv- ing decision-making behavior and related environmental factors affecting driving decision-making, a single-track variable speed decision model based on fuzzy support vector machine has been proposed in combination with fuzzy theory and soft interval support vector machine theory, which aims to provide drivers with decision-making suggestions for the current situation and to efficiently help drivers to develop a safe driving strategy. The simulation results show that compared with the traditional support vector machine, the proposed model has higher decision accuracy and relatively good expansibility suitable to provide drivers reasonable and safe recommendations.
出处
《计算机技术与发展》
2017年第7期190-193,199,共5页
Computer Technology and Development
基金
国家自然科学基金资助项目(61501251
61373137
61071167)
江苏省普通高校研究生科研创新计划资助项目(KYZZ15_0236)
南京邮电大学引进人才科研启动基金资助项目(NY214191)
关键词
单向变速决策模型
模糊理论
软间隔支持向量机
隶属函数
single-track variable speed decision model
fuzzy theory
soft interval support vector machine
membership functions
作者简介
尤永健（1995-），男，研究方向为机器学习及最优化理论；
李雷，博士，教授，研究方向为机器学习及最优化理论。
