机动车驾驶人驾驶行为不确定性建模与仿真

发布时间：2018-05-19 06:32

本文选题：驾驶行为操纵模式 + 主成分分析法(PCA)　；参考：《合肥工业大学》2015年硕士论文

【摘要】：随着现代社会的发展,车辆已经成为人们日常所不可或缺的交通工具,然而,随着汽车保有量的增加,相应的问题也与之而来,及交通事故发生率的增加,而在交通事故中,驾驶员这一因素的影响又至关重要,所以需要提高驾驶员的驾驶能力。基于以上因素,出现了驾驶辅助系统,他可以为驾驶员提供辅助和支持,在潜在危险出现时,可以为驾驶员提供报警信息以及对车辆的辅助控制。但是由于驾驶员个体之间的驾驶特性完全不一样,座椅给辅助系统的设计造成了较大的困难,如何使系统在保证安全的前提下去适应驾驶员差异,成为系统设计的一个问题。根据以上问题的出现,本文中主要的研究内容如下： (1)通过采集并分析驾驶行为与驾驶操作动作模式的数据,利用主成分分析法研究驾驶行为及其对应的驾驶操纵动作模式之间的关系； (2)根据个体驾驶人完成一个具体驾驶行为的驾驶操纵动作具有一定的内聚性、时序性和个性化的机理,研究构建基于有向图的驾驶人驾驶行为操作模式建模方法； (3)根据前面利用主成分分析法提取出的对于各个驾驶行为占主要影响的驾驶操作动作特征数据,设计针对不同特征主成分和各个特征向量组合的神经网络算法,可以反向实现个性化驾驶行为的识别,进一步为利用驾驶行为及驾驶操纵动作模式设计先进的驾驶辅助系统提供技术支撑。研究的结果表明： (1)主成分分析法可以确定出对于特定的驾驶行为影响较为关键的驾驶操作动作； (2)可以利用MATLAB仿真软件构建基于有向图模式的驾驶行为操作模式,可以用于建立机动车驾驶人驾驶行为操纵模型,有利于自适应系统的建立； (3)Fuzzy_ARTMAP神经网络驾驶行为识别算法具有较高的准确率以及较低的虚警率,该方法可以作为驾驶行为识别中一个有效可行的解决方案。
[Abstract]:With the development of modern society, vehicles have become an indispensable means of transportation for people. However, with the increase of vehicle ownership, the corresponding problems also arise, and the incidence of traffic accidents increases, and in traffic accidents, The driver's influence is also very important, so it is necessary to improve the driver's driving ability. Based on the above factors, there is a driving assistance system, which can provide assistance and support for the driver, and can provide alarm information and auxiliary control to the vehicle when the potential danger arises. However, since the driving characteristics of individual drivers are completely different, it is difficult for the seat to design the auxiliary system. How to make the system adapt to the drivers' differences on the premise of ensuring safety has become a problem in the system design. According to the above problems, the main contents of this paper are as follows: 1) by collecting and analyzing the data of driving behavior and driving operation mode, the relationship between driving behavior and its corresponding driving operation mode is studied by principal component analysis. 2) according to the mechanism that individual driver has certain cohesion, timing and individuation to complete a specific driving behavior, the modeling method of driving behavior operation mode based on directed graph is studied. (3) according to the characteristic data of driving operation, which is extracted by principal component analysis (PCA), a neural network algorithm is designed for different feature principal components and combination of each characteristic vector. It can realize the recognition of individual driving behavior in reverse direction, and provide technical support for the design of advanced driving assistant system by using driving behavior and driving manipulation mode. The results of the study show that: 1) Principal component analysis (PCA) can determine the key driving actions that affect the specific driving behavior. 2) the driving behavior operation mode based on directed graph mode can be constructed by using MATLAB simulation software, and it can be used to establish the driving behavior control model of motor vehicle, which is beneficial to the establishment of adaptive system. FuzzyARMAP neural network has high accuracy and low false alarm rate. This method can be used as an effective and feasible solution in driving behavior recognition.
【学位授予单位】：合肥工业大学
【学位级别】：硕士
【学位授予年份】：2015
【分类号】：U491.25;U463.6

【参考文献】