本文關(guān)鍵詞： 交通建模 模糊決策 動態(tài)軌跡規(guī)劃 交通仿真系統(tǒng)　出處：《吉林大學(xué)》2015年碩士論文　論文類型：學(xué)位論文

【摘要】：與傳統(tǒng)汽車技術(shù)研發(fā)不同，汽車智能化系統(tǒng)與汽車行駛環(huán)境、特別是交通環(huán)境高度相關(guān)。交通和交通車輛對智能汽車或汽車智能行駛過程的環(huán)境傳感感知、軌跡規(guī)劃和自動控制等具有十分重要的影響，是關(guān)乎汽車智能行駛安全和通行效率的關(guān)鍵因素。當(dāng)前基于試驗場地的實車試驗仍然是汽車智能化技術(shù)和產(chǎn)品開發(fā)、測試和驗證的主流平臺，但實車試驗往往成本高、周期長、可重復(fù)性差且往往具有較高的危險性。雖然模擬仿真技術(shù)越來越成為了汽車產(chǎn)品研發(fā)的重要技術(shù)手段，但現(xiàn)有汽車模擬仿真平臺多側(cè)重于車輛本身的性能問題，而較少關(guān)注對汽車行駛環(huán)境和環(huán)境傳感的模擬，因此難于滿足對汽車智能化技術(shù)的虛擬研發(fā)需求。因此研發(fā)面向汽車智能化仿真的交通模擬仿真技術(shù)和仿真平臺就具有十分重要的意義。 不同于基于多剛體力學(xué)的傳統(tǒng)汽車模擬仿真技術(shù)，也不同于基于流體力學(xué)的傳統(tǒng)交通模擬仿真技術(shù)，面向汽車智能化的交通仿真技術(shù)是一個較新的研究領(lǐng)域，其目的在于為車載環(huán)境傳感與感知、軌跡規(guī)劃和智能決策、智能或無人駕駛等提供具有較高置信度的、且能與智能汽車形成相互交互的實時交通環(huán)境，因此充滿著技術(shù)挑戰(zhàn)。 隨著基于計算機圖形學(xué)的汽車行駛環(huán)境三維建模技術(shù)的發(fā)展，隨著車載像機和雷達等環(huán)境傳感器的模擬仿真技術(shù)的發(fā)展，對面向智能汽車行駛的交通環(huán)境模擬提出了新的要求，包括具有交通流的隨機性特征、具有車輛動力學(xué)的特征和具有主動控制的特征等，使之能與智能汽車形成相互交互。此外，模型的置信度和高效實時性也是交通車輛和環(huán)境模擬仿真的關(guān)鍵要求，不僅可以彌補實車試驗的不足、部分替代以減少實車試驗，也是提高智能汽車技術(shù)與產(chǎn)品研發(fā)質(zhì)量和效率的關(guān)鍵技術(shù)。 本文通過對現(xiàn)有研究現(xiàn)狀的深入調(diào)研，以汽車智能化仿真需求為導(dǎo)向，對面向汽車智能化仿真的交通模擬方法進行了研究。在此基礎(chǔ)上開發(fā)了具有不同宏觀特征的交通流模型和微觀交通車輛模型，使交通車輛具有自主駕駛和行為決策等能力，實現(xiàn)了交通車輛之間，交通車與智能汽車之間的交互行為，構(gòu)建出復(fù)雜而又真實的交通試驗環(huán)境。 本文的主要工作和成果如下： 1．開展了對交通和交通車輛模擬方法的研究，并據(jù)此設(shè)計了其模型。該模型包括發(fā)車模型和智能交通車輛模型。發(fā)車模型包括交通流宏觀參數(shù)，以產(chǎn)生符合需求的交通流。智能交通車輛模型能夠根據(jù)道路模型以及周圍交通車輛信息等自主行駛。 2.開展了對車輛跟馳理論的研究；在擴展型IDM模型基礎(chǔ)上，通過改進加速度調(diào)整項解決了其在車輛重啟動方面的不足。 3.研究了智能交通車輛的行為決策方法，建立了基于模糊推理的換道決策模型。提出以駕駛員不滿積累程度為換道決策行為產(chǎn)生的客觀評價指標(biāo)。該模型考慮了換道時影響駕駛員換道愿望的主客觀因素，分別以換道收益和駕駛員魯莽程度作為評價指標(biāo)，另外以潛在沖突可能性評價換道可行性。采用模糊邏輯的方法將影響換道決策的各個因素聯(lián)系起來，有效地模擬了真實駕駛員的行為決策過程，使得智能交通車輛的行為決策表現(xiàn)具有較高的置信度，充分反映了實際交通流的動態(tài)不確定性，提高了交通模擬的置信度。 4.針對智能交通車輛的換道過程，提出了基于多目標(biāo)約束的動態(tài)軌跡規(guī)劃方法。該方法通過二自由度車輛動力學(xué)模型生成軌跡，不僅考慮了車輛的動力學(xué)特性，而且計算效率高，實時性好。該軌跡可以根據(jù)安全和效率的不同需求進行最優(yōu)化調(diào)整，，結(jié)合交通車軌跡預(yù)估，能夠在復(fù)雜的道路交通環(huán)境中快速地規(guī)劃出一條兼顧安全性、舒適性、操縱穩(wěn)定性和駕駛效率等多目標(biāo)約束的軌跡，確保了對動態(tài)交通環(huán)境的快速響應(yīng)，使得智能交通車輛在道路行駛中具有較高的真實度。 5.在交通建模理論研究的基礎(chǔ)上，設(shè)計并編程實現(xiàn)了面向汽車智能化仿真的交通模擬系統(tǒng)。設(shè)計了輔助的道路結(jié)構(gòu)和依托于道路結(jié)構(gòu)的車輛管理體系。通過測試證明了理論模型和系統(tǒng)的有效性。
[Abstract]:Different from the traditional automobile technology development, intelligent vehicle system and road environment, especially traffic environment are highly correlated. The traffic and transport vehicles for environmental sensing perception of intelligent vehicle or vehicle intelligent, has important implications for trajectory planning and control, is a key factor related to the intelligent vehicle driving safety and traffic efficiency based on the real vehicle test. The test site is still the intelligent vehicle technology and product development, testing and validation of the mainstream platform, but the real vehicle tests are often high cost, long cycle, risk of poor repeatability and often have high. Although analog simulation technology becomes an important technical means of automotive product development however, the problem that the existing automobile performance simulation platform mainly focus on the vehicle itself, but pay less attention to the vehicle simulation environment and environmental sensing, so difficult In order to meet the demand of virtual research and development for automotive intelligent technology, it is very important to develop vehicle oriented intelligent simulation technology and simulation platform.
Different from the simulation of multi rigid body mechanics is different from the traditional car based on the fluid mechanics of traditional traffic simulation technology based on traffic simulation technology of the intelligent vehicle is a relatively new field of study, its purpose is to provide the vehicle environment sensing and perception, path planning and intelligent decision-making, intelligent or unmanned etc. provide high reliability, real-time traffic environment and the formation of mutual interaction and intelligent vehicle, so full of technical challenges.
With the development of 3D modeling technology of driving environment of computer graphics based on the car, with the development of vehicle like machine simulation technology and radar sensor environment, new requirements oriented intelligent vehicle traffic environment simulation is proposed, including random characteristics of traffic flow, with characteristics of vehicle dynamics and characteristic of active control so, to form mutual interaction and intelligent vehicle. In addition, the key requirement of simulation confidence and efficient real-time is traffic and environment model, can not only make up the inadequacy of the real vehicle test, in order to reduce the partial replacement of real vehicle tests, but also improve the key technology of intelligent vehicle technology and product development quality and efficiency.
Through in-depth research on the status quo of the existing research, the intelligent vehicle simulation oriented, oriented intelligent simulation of automobile traffic simulation method is studied. On the basis of the development of the model and the micro vehicle traffic flow model with different macroscopic characteristics, so that the vehicle has autonomous driving and behavior decision. To achieve the traffic between vehicles, interactions between traffic and intelligent vehicle traffic, build the experimental environment is complicated and real.
The main work and results of this article are as follows:
1. research on simulation method for traffic and traffic vehicles, and based on the model. The model includes the grid model and intelligent vehicle model. The grid model including the macroscopic traffic flow parameters, to meet the needs of the traffic flow. The intelligent transportation vehicle model according to the road and the surrounding traffic information model of autonomous driving.
2., vehicle following theory has been studied. Based on the extended IDM model, it has solved its shortcomings in vehicle restart by improving the acceleration adjustment.
3.鐮旂┒浜嗘櫤鑳戒氦閫氳濺杈嗙殑琛屼負(fù)鍐崇瓥鏂規(guī)硶,寤虹珛浜嗗熀浜庢ā緋婃帹鐞嗙殑鎹㈤亾鍐崇瓥妯″瀷.鎻愬嚭浠ラ┚椹跺憳涓嶆弧縐瘡紼嬪害涓烘崲閬撳喅絳栬涓轟駭鐢熺殑瀹㈣璇勪環(huán)鎸囨爣.璇ユā鍨嬭

本文編號：1474501