本文選題：多目標(biāo)優(yōu)化　切入點(diǎn)：過(guò)飽和干線　出處：《鄭州大學(xué)》2017年碩士論文　論文類型：學(xué)位論文

【摘要】：城市交通擁堵問(wèn)題日趨突出,干線在車流高峰時(shí)期處于過(guò)飽和狀態(tài)已成常態(tài),交叉口信號(hào)配時(shí)直接控制車輛通行,進(jìn)而影響干線交通效率,如何優(yōu)化過(guò)飽和交通干線的信號(hào)配時(shí)方案成為亟待研究的問(wèn)題。本文針對(duì)城市交通過(guò)飽和狀態(tài)下干線信號(hào)配時(shí)優(yōu)化問(wèn)題,分析當(dāng)前已有方法的異同及適用情況,確定結(jié)合微觀交通仿真實(shí)現(xiàn)信號(hào)配時(shí)方案多目標(biāo)多變量?jī)?yōu)化。主要工作如下:(1)分析柵格地圖特點(diǎn),采用圖像處理方法完成干線路網(wǎng)自動(dòng)化提取與構(gòu)建,結(jié)合城市交通特點(diǎn)研究車輛跟馳與換道模型,自主構(gòu)建微觀仿真環(huán)境以評(píng)價(jià)信號(hào)配時(shí)方案;(2)選用快速非支配遺傳算法NSGAⅡ?qū)崿F(xiàn)多目標(biāo)優(yōu)化,分析該算法中重復(fù)個(gè)體對(duì)小種群進(jìn)化結(jié)果的影響,采用預(yù)選擇策略剔除個(gè)體并提出快速重復(fù)檢測(cè)方法,提高了算法性能;(3)基于交通信號(hào)控制相關(guān)理論,分析過(guò)飽和狀態(tài)下不同優(yōu)化參數(shù)及目標(biāo)對(duì)交通效率的影響,提出以綠信比、相序、相位差和周期為優(yōu)化參數(shù),車輛平均時(shí)延、系統(tǒng)平均排隊(duì)-車道長(zhǎng)度比和系統(tǒng)通行能力為優(yōu)化目標(biāo)的交通信號(hào)優(yōu)化模型,并結(jié)合仿真環(huán)境和改進(jìn)算法完成對(duì)干線各交叉口信號(hào)配時(shí)方案的優(yōu)化。利用采集的交通數(shù)據(jù)對(duì)由三個(gè)交叉口組成的干線進(jìn)行實(shí)例驗(yàn)證,實(shí)驗(yàn)結(jié)果表明在過(guò)飽和狀態(tài)下,本文提出的信號(hào)優(yōu)化方法與其他方法相比不僅有效控制車輛排隊(duì)長(zhǎng)度,均衡車輛分布,同時(shí)在系統(tǒng)通行能力,車均時(shí)延方面表現(xiàn)更佳。
[Abstract]:The problem of urban traffic congestion is becoming more and more prominent, and it is normal for the trunk lines to be supersaturated during the rush hour of traffic flow. The signalling timing of intersections directly controls the traffic flow, thus affecting the traffic efficiency of the trunk lines. How to optimize the signal timing scheme of supersaturated traffic trunk lines is an urgent problem to be studied. In this paper, the similarities and differences of the existing methods and their application are analyzed in view of the problem of signal timing optimization of trunk lines under the condition of city crossing saturation. The main work is as follows: analyzing the characteristics of grid map, using image processing method to complete the automatic extraction and construction of trunk road network. According to the characteristics of urban traffic, the following and changing models of vehicles are studied, and the microscopic simulation environment is built independently to evaluate the signal timing scheme.) the fast non-dominant genetic algorithm (NSGA 鈪，

本文編號(hào)：1641787