摘要
个体受限于认知能力和逻辑推理能力的限制,在出行决策过程中很难做到完全理性.本文以出发时间选择为例,在有限理性行为假设基础上,引入空间知识获取、学习及认知更新和方案搜索等关键行为要素,构建有限理性下的出行决策过程理论框架.融合RP和SP调查方法,设计出发时间选择行为意向调查方案.研究个体知识的表达方式,应用贝叶斯学习理论完成认知更新.定义搜索成本和收益函数,利用调查数据分别提取基于PART和RIPPER算法的出发时间启发式搜索规则和决策规则.结果表明,有限理性下个体出发时间选择行为存在感知阈值,而并非寻求全局最优解.
Subject to such limitations as cognitive ability and logical reasoning ability, it is difficult for individual to be perfectly rational in the travel decision-making process. Taking departure time choice as an example, key behavior factors such as spatial knowledge acquisition, learning, cognition update and solution search are introduced. The theoretical framework of travel decision-making process is built based on bounded rationality. Departure time behavioral intention survey program is designed by integration of RP and SP survey methods. Individual' s knowledge representation is studied, and cognitive update is completed by using Bayesian learning theory. Functions of search cost and search gain are defined. By using survey data, departure time heuristic search rules and decision rules are derived based on PART and RIPPER algorithm separately. The results show that there exist perception threshold in individual' s departure time choice behavior under bounded rationality, rather than seeking global optimal solution.
出处
《交通运输系统工程与信息》
EI
CSCD
北大核心
2016年第1期135-141,共7页
Journal of Transportation Systems Engineering and Information Technology
基金
国家自然科学青年基金项目(71403096)
江苏省教育厅高校哲学社会科学基金项目(2013SJD630005)
住房和城乡建设部科学技术计划项目(2013-k5-4)~~