智能车路径跟踪控制算法的简述
2020-02-25杨浩
摘 要:智能车主要分为路径规划、路径跟踪、自动泊车三大部分。路径规划主要研究车辆的避障问题,路径跟踪主要研究车辆跟随期望路径的有效性,自动泊车主要分析车辆在有限的几何空间内将车辆泊到指定的空间位置。其中路径跟踪是其核心部分,根据研究方法的不同,主要分为“预瞄跟随模型”和“智能控制模型”。文章根据预瞄点的不同,主要分析单点预瞄模型、两点预瞄模型、路程预瞄模型。根据智能控制方法的不同,主要分析模糊逻辑控制驾驶员模型、神经网络控制驾驶员模型、模型预测控制驾驶员模型。主题词:智能车;预瞄跟随模型;智能控制模型中图分类号:U471 文献标识码:A 文章编号:1671-7988(2020)01-49-02
Abstract: Smart cars are mainly divided into three parts: path planning, path tracking, and automatic parking. Path planning mainly studies the obstacle avoidance problem of vehicles. Path tracking mainly studies the effectiveness of vehicles following the expected path. Automatic parking mainly analyzes vehicles parking vehicles in a limited geometric space to a specified spatial position. Path tracking is its core part. According to different research methods, it is mainly divided into “pre-targeting model” and “intelligent control model”.According to the different pre-pointing points, this paper mainly analyzes the single-point preview model, the two-point preview model, and the route preview model. According to the different intelligent control methods, the fuzzy logic control driver model, the neural network control driver model, and the model predictive control driver model are mainly analyzed.Keywords: Smart car vehicle; Preview follower model; Intelligent control modelCLC NO.: U471 Document Code: A Article ID: 1671-7988(2020)01-49-02
前言
駕驶员模型在人-车-路闭环系统中有着重要作用。为了建立更逼真的驾驶员模型,充分考虑驾驶员的前视作用。文献[1]建立了一种单点预瞄的驾驶员模型。文献[2]建立了一种两点预瞄的驾驶员模型。文献[3]建立了更一般的预瞄形式即路程预瞄驾驶员模型。随着控制理论的发展,有关学者将智能控制方法引入到了驾驶员模型,建立了一类具有推理思维的驾驶员模型。文献[4]结合模糊逻辑控制方法具有模拟人的推理过程的优势,建立了一种模糊逻辑智能控制驾驶员模型。文献[5]基于样本的数据,建立了一种神经网络驾驶员模型。文献[6]结合模型预测控制算法具有预测和滚动优化的优势,建立了一种基于模型预测控制的驾驶员模型。……
