交通运输研究
交通運輸研究
교통운수연구
Transportation Standardization
2015年
4期
49-55
,共7页
交通诱导%累积自学习机制%模糊博弈%驾驶员路径选择%有限理性
交通誘導%纍積自學習機製%模糊博弈%駕駛員路徑選擇%有限理性
교통유도%루적자학습궤제%모호박혁%가사원로경선택%유한이성
traffic guidance%cumulative self-learning mechanism%fuzzy game%drivers′route selec-tion%bounded rationality
为弥补已有驾驶员路径选择博弈模型将驾驶员视为完全理性的不足，探求无诱导信息情况下路网交通流临界状态，将驾驶员视为有限理性，其依赖累积时间感受收益做出下一次的路径选择策略，并以驾驶员的行程时间感受作为决策收益建立了基于累积自学习机制的无诱导信息驾驶员路径选择博弈模型。利用该模型，分析了驾驶员路径选择行为对路网交通流的影响，并通过仿真验证得出了不同初始状态下的模型博弈平衡结果。仿真结果表明：博弈平衡状态与路网车流总量及初始流量分配比例密切相关。当路网车流总量小于或接近路网总通行能力时，不发布诱导信息，路网的交通流分布达到稳定平衡，路网通行能力利用率较高；当路网车流总量远大于路网总通行能力时，不发布诱导信息，路网交通流分布会形成峰谷平衡，不能有效利用路网通行能力，应采取相应的交通管理措施。
為瀰補已有駕駛員路徑選擇博弈模型將駕駛員視為完全理性的不足，探求無誘導信息情況下路網交通流臨界狀態，將駕駛員視為有限理性，其依賴纍積時間感受收益做齣下一次的路徑選擇策略，併以駕駛員的行程時間感受作為決策收益建立瞭基于纍積自學習機製的無誘導信息駕駛員路徑選擇博弈模型。利用該模型，分析瞭駕駛員路徑選擇行為對路網交通流的影響，併通過倣真驗證得齣瞭不同初始狀態下的模型博弈平衡結果。倣真結果錶明：博弈平衡狀態與路網車流總量及初始流量分配比例密切相關。噹路網車流總量小于或接近路網總通行能力時，不髮佈誘導信息，路網的交通流分佈達到穩定平衡，路網通行能力利用率較高；噹路網車流總量遠大于路網總通行能力時，不髮佈誘導信息，路網交通流分佈會形成峰穀平衡，不能有效利用路網通行能力，應採取相應的交通管理措施。
위미보이유가사원로경선택박혁모형장가사원시위완전이성적불족，탐구무유도신식정황하로망교통류림계상태，장가사원시위유한이성，기의뢰루적시간감수수익주출하일차적로경선택책략，병이가사원적행정시간감수작위결책수익건립료기우루적자학습궤제적무유도신식가사원로경선택박혁모형。이용해모형，분석료가사원로경선택행위대로망교통류적영향，병통과방진험증득출료불동초시상태하적모형박혁평형결과。방진결과표명：박혁평형상태여로망차류총량급초시류량분배비례밀절상관。당로망차류총량소우혹접근로망총통행능력시，불발포유도신식，로망적교통류분포체도은정평형，로망통행능력이용솔교고；당로망차류총량원대우로망총통행능력시，불발포유도신식，로망교통류분포회형성봉곡평형，불능유효이용로망통행능력，응채취상응적교통관리조시。
In order to make up the shortage of regarding the drivers as totally rational by the existing game theory models of driver′s route selection, and search the fettle of network traffic flow without induc?ing information, the drivers′rationality was deemed bounded, so he or she made the selection decision of next path according to the cumulative feeling gains of time. The driver′s feeling of travel time was consid?ered as the gains of decision, then the model of the driver′s routing selection without inducing informa?tion was established based on cumulative self-learning mechanism, and the impact of the driver′s route choice behavior on road network traffic flow was analyzed. Finally the equilibrium results of fuzzy game on different initial states were obtained by the simulation. The simulation results show that the balance of fuzzy game is closely related to the total amount of road network traffic and initiatory flow distribution ratio. When the total amount of road network traffic is less than or close to the capacity of network, the distribution of network traffic flow reaches a stable equilibrium and the utilization rate of the capacity of network is higher without inducing information. When the total amount of road network traffic is over the capacity of network, the distribution of network traffic flow forms a peak balance, the capacity of network can′t be used effectively, and appropriate traffic management measures should be taken.
