第1 章 智能體運(yùn)動(dòng)規(guī)劃研究概述·······································································（1）
1.1 研究背景···························································································（1）
1.2 研究現(xiàn)狀···························································································（1）
1.3 研究思路···························································································（8）
第2 章 運(yùn)動(dòng)規(guī)劃中的智能體多層次行為模型框架設(shè)計(jì)·········································.（11）
2.1 環(huán)境建模方法··················································································.（11）
2.1.1 三角剖分法···········································································.（11）
2.1.2 可視圖方法···········································································.（12）
2.1.3 Voronoi 圖法··········································································.（12）
2.1.4 隨機(jī)采樣法···········································································.（13）
2.2 運(yùn)動(dòng)規(guī)劃方法··················································································.（16）
2.2.1 全局運(yùn)動(dòng)規(guī)劃········································································.（16）
2.2.2 局部運(yùn)動(dòng)規(guī)劃········································································.（20）
2.3 智能體多層次行為模型設(shè)計(jì)································································.（23）
2.3.1 智能體介紹···········································································.（23）
2.3.2 現(xiàn)有智能體行為模型框架·························································.（24）
2.3.3 智能體多層次行為模型框架······················································.（27）
第3 章 基于多信息域、多分辨率場(chǎng)景描述模型的多層次運(yùn)動(dòng)規(guī)劃算法研究··············.（31）
3.1 相關(guān)工作························································································.（31）
3.2 層次化運(yùn)動(dòng)規(guī)劃算法總體思路·····························································.（34）
3.2.1 問題描述··············································································.（34）
3.2.2 總體思路··············································································.（35）
3.3 全局層次上的多樣化引導(dǎo)路徑生成·······················································.（36）
3.3.1 高層時(shí)空約束的表示·······························································.（36）
3.3.2 全局概率路徑圖的生成····························································.（38）
3.3.3 全局多樣化引導(dǎo)路徑的生成······················································.（39）
3.4 局部層次上的多樣化運(yùn)動(dòng)路徑生成·······················································.（41）
3.4.1 局部高分辨率概率路徑圖的生成················································.（41）
3.4.2 局部多樣化運(yùn)動(dòng)路徑的生成······················································.（42）
3.5 動(dòng)態(tài)障礙情況下的運(yùn)動(dòng)路徑重新規(guī)劃問題··············································.（43）
3.5.1 運(yùn)動(dòng)路徑的重新規(guī)劃·······························································.（43）
3.5.2 概率路徑圖的動(dòng)態(tài)更新····························································.（44）
3.6 仿真實(shí)驗(yàn)與結(jié)果分析·········································································.（45）
第4 章 多智能體避碰行為研究·······································································.（48）
4.1 相關(guān)工作························································································.（48）
4.2 避碰行為的基本概念與問題描述··························································.（50）
4.2.1 基本概念··············································································.（50）
4.2.2 問題描述··············································································.（51）
4.3 反應(yīng)式避碰行為模型·········································································.（52）
4.4 基于最小代價(jià)原則的預(yù)測(cè)式避碰行為模型··············································.（57）
4.4.1 最小代價(jià)原則········································································.（57）
4.4.2 基于最小代價(jià)原則的預(yù)測(cè)式避碰行為建模····································.（58）
4.5 仿真實(shí)驗(yàn)與結(jié)果分析·········································································.（63）
第5 章 基于增廣物理仿真的運(yùn)動(dòng)規(guī)劃多任務(wù)約束建模分析與優(yōu)化求解····················.（67）
5.1 相關(guān)工作························································································.（67）
5.2 問題描述························································································.（70）
5.3 控制圍欄函數(shù)基礎(chǔ)背景知識(shí)································································.（70）
5.4 基于控制圍欄函數(shù)的多任務(wù)約束統(tǒng)一描述與運(yùn)動(dòng)路徑動(dòng)力學(xué)物理仿真優(yōu)化生成···.（71）
5.4.1 基于控制圍欄函數(shù)的多任務(wù)約束統(tǒng)一描述····································.（72）
5.4.2 增廣物理仿真框架下的智能體運(yùn)動(dòng)路徑優(yōu)化生成···························.（74）
5.5 仿真實(shí)驗(yàn)與結(jié)果分析·········································································.（77）
5.5.1 可生成多樣化行為的智能體運(yùn)動(dòng)路徑優(yōu)化計(jì)算······························.（77）
5.5.2 自動(dòng)駕駛中自適應(yīng)巡航運(yùn)動(dòng)路徑優(yōu)化計(jì)算····································.（81）
5.6 多智能體運(yùn)動(dòng)規(guī)劃仿真演示軟件設(shè)計(jì)開發(fā)··············································.（83）
5.6.1 演示軟件設(shè)計(jì)開發(fā)··································································.（83）
5.6.2 多智能體運(yùn)動(dòng)路徑生成測(cè)試······················································.（85）
第6 章 多任務(wù)約束時(shí)空融合處理機(jī)制研究························································.（87）
6.1 相關(guān)工作························································································.（87）
6.2 具有速度和加速度約束的多智能體時(shí)間最優(yōu)軌跡規(guī)劃·······························.（88）
6.2.1 問題描述··············································································.（88）
6.2.2 算法思路及流程·····································································.（92）
6.2.3 可達(dá)集與可達(dá)速度區(qū)間的計(jì)算···················································.（95）
6.2.4 利用VIP 算法計(jì)算時(shí)間最優(yōu)軌跡················································.（97）
6.2.5 仿真實(shí)驗(yàn)與結(jié)果分析·······························································.（99）
6.3 基于控制圍欄函數(shù)描述的多任務(wù)約束時(shí)空融合處理··································（107）
6.3.1 同時(shí)段下多任務(wù)約束時(shí)序沖突的自動(dòng)優(yōu)化處理······························（107）
6.3.2 仿真實(shí)驗(yàn)與結(jié)果分析·······························································（111）
第7 章 滿足時(shí)空約束要求的多智能體運(yùn)動(dòng)規(guī)劃在城市維穩(wěn)處突行動(dòng)仿真場(chǎng)景下的應(yīng)用····.（115）
7.1 應(yīng)用場(chǎng)景描述··················································································（115）
7.2 可視化仿真·····················································································（117）
7.2.1 仿真參數(shù)設(shè)置········································································（117）
7.2.2 仿真結(jié)果演示········································································（118）
第8 章 深度強(qiáng)化學(xué)習(xí)在運(yùn)動(dòng)規(guī)劃中的應(yīng)用探索研究············································（124）
8.1 相關(guān)工作························································································（124）
8.2 深度強(qiáng)化學(xué)習(xí)算法探索效率提升策略研究··············································（126）
8.2.1 基本思路··············································································（127）
8.2.2 遺傳算法和DDPG 算法相結(jié)合的DRL 算法探索效率提升策略··········（129）
8.2.3 仿真實(shí)驗(yàn)與結(jié)果分析·······························································（132）
8.3 基于深度強(qiáng)化學(xué)習(xí)的多智能體避碰行為生成···········································（134）
8.3.1 問題描述與基本思路·······························································（134）
8.3.2 仿真實(shí)驗(yàn)與結(jié)果分析·······························································（135）
8.4 下一步展望·····················································································（136）
參考文獻(xiàn)······································································································（137）