1、 题目：题目：基于事件相关电位的测谎方法研究 学学 院院 信息科学与工程学院 专专 业业 自动化 年年 级级 2010 级 学学 号号 姓姓 名名 指导老师指导老师 2014 年年 5 月月 摘摘 要要 本研究主要利用P300电位作为辨别测谎的重要指标。P300测谎与任务相关性、刺激性 质、注意、记忆、电生理干扰等影响因素有关。利用P300测谎的目的是对单一个体进行其 是否在说谎的判断，本课题研究主要关注单个被试内的探测刺激和无关刺激的P300是否存 在显著差异。本文采用个人相关信息为刺激内容，以被试真实的相关信息作为目标刺激， 例如被试的存钱银行名称等， 设计实验范式， 将实验采集到的脑电信
2、号作为主要研究对象。 P300电位的数据处理分为三步：预处理，特征提取和分类。将采集到的数据通过汉明窗低 通滤波器滤波，选择导联和时间窗进行降维，然后利用Fisher线性判别进行分类识别。对 分类结果进行分析得出：正确率与训练量的多少、刺激重复次数、导联选取个数均有关系。 关键词关键词：事件相关电位,P300 电位,Fisher 线性判别 ABSTRACTABSTRACT This research mainly uses the P300 potential as an important index to distinguish the polygraph. The relevant fa
3、ctors of P300 polygraph and task dependencies, stimulating properties, attention, memory, electrophysiological disturbance. Using P300 lie to the single individual is lying judgment, this research mainly focus on the detection of single subjects and tn different stimuli P300 whether there are signif
4、icant differences. This paper uses personal information related to stimulate the content, related to the information is true as target stimuli, such as the bank name, design experimental paradigm, the EEG signal is acquired from experiments as the main object of study. Data processing of P300 potent
5、ial is divided into three steps: preprocessing, feature extraction and classification. The data will be collected by the Hamming window low-pass filter, select lead and time window to reduce the dimension, and then use the Fisher linear classification discrimination. The classification results are obtained: the correct rate and amount of training, stimulus repetition, lead the chosen number of. Keywords: event related potential，P300，Fisher linear discriminant 目目 录录 第一章第一章 绪论 绪论 1 1 1.1 引言
