Adaptive EEG-Based Alertness Estimation System by Using ICA-Based Fuzzy Neural Networks

Drivers' fatigue has been implicated as a causal factor in many accidents. The development of human cognitive state monitoring system for the drivers to prevent accidents behind the steering wheel has become a major focus in the field of safety driving. It requires a technique that can continuously monitor and estimate the alertness level of drivers. The difficulties in developing such a system are lack of significant index for detecting drowsiness and the interference of the complicated noise in a realistic and dynamic driving environment. An adaptive alertness estimation methodology based on electroencephalogram, power spectrum analysis, independent component analysis (ICA), and fuzzy neural network (FNNs) models is proposed in this paper for continuously monitoring driver's drowsiness level with concurrent changes in the alertness level. A novel adaptive feature selection mechanism is developed for automatically selecting effective frequency bands of ICA components for realizing an on-line alertness monitoring system based on the correlation analysis between the time-frequency power spectra of ICA components and the driving errors defined as the deviation between the center of the vehicle and the cruising lane in the virtual-reality driving environment. The mechanism also provides effective and efficient features that can be fed into ICA-mixture-model-based self-constructing FNN to indirectly estimate driver's drowsiness level expressed by approximately and predicting the driving error     

基于脑电图识别结合操纵特征的驾驶疲劳检测分析

交警部门在进行道路安全管理时,对疲劳驾驶的人员进行有效的驾驶疲劳检测,是辨别疲劳驾驶人员前提与基础。本文基于脑电图识别结合操纵特征为切入点,通过选取的样本进行驾驶疲劳实验,将脑电图识别与车辆操纵特性相结合来检测驾驶员的疲劳状态。     

基于脑电识别与车辆操纵特征的驾驶疲劳检测分析

本文以脑电识别与车辆操纵特征为切入点,通过模拟疲劳驾驶实验,将脑电识别与车辆操纵特性相结合来检测驾驶员的疲劳状态.通过对脑电信号的S变换分析,发现不同驾驶时刻其变换时频谱图存在显著差异,可用来区分驾驶过程中驾驶员的精神状态,结合车辆操纵特征参数,得到操纵特征与疲劳状态的关系,为脑电识别与操纵特征的驾驶疲劳检测的有效性提供一定的理论和实验基础.     

Seat Belt Vibration as a Stimulating Device for Awakening Drivers

This paper presents a safety driving system that uses a seat belt vibration as a stimulating device for awakening drivers. The vibration stimulus was composed of pulsation tension, which was applied by the seat belt motor retractor. Magnitude, duration, and repetition rate of the additional tension were the major parameters that determined the awakening effect of the stimulus. We constructed a driving simulator, which was able to induce driver's drowsiness. In the experiments using the driving simulator, the driver's drowsiness was detected by changes in the driver's eye movements measured by electrooculography (EOG) and/or changes in facial expression of the driver monitored by the examiners through a video camera, subjective evaluation, and lane deviation. Exerting additional tension of 130 N for 3 cycles at duration and interval of 100 ms was the most effective pattern for awakening the driver without causing discomfort.     

Driver Behavior and Situation Aware Brake Assistance for Intelligent Vehicles

This paper deals with the development of Human-Centric Intelligent Driver Assistance Systems. Rear-end collisions account for a large portion of traffic accidents. To help mitigate this problem, predictive braking systems and adaptive cruise control systems have been developed. However, these types of systems usually rely solely on the vehicle and vehicle surround sensors, either ignoring the human component of driving or learning the driver's control behavior using only these sensors. As with all human-computer interfaces, this has the potential to work against the driver, distract the driver further, or even annoy the driver so that the driver ignores or disables the system. It is, therefore, important to directly take the driver's intended actions into account when designing a driver assistance system. By using a probabilistic model for the system, warnings and preventative measures can be constructed based on varying levels of situational severity and driver attentiveness and intent. The research is based upon carefully conducted experimental trials involving a human subjects driving in natural manner and on typical freeways in the USA. The experiments, designed by inputs from cognitive scientist, were conducted in a specially designed instrumented vehicle to record important cues associated with driver's behavior, vehicle state, and vehicle surround in a synchronized manner. Quantitative results and analysis of the experimental trials are presented to show the feasibility and promise of this framework to predict the driver's intent to brake, the need for braking given the current situation, and at what level the driver should be warned     

EEG-based assessment of driver cognitive responses in a dynamic virtual-reality driving environment

Accidents caused by errors and failures in human performance among traffic fatalities have a high death rate and become an important issue in public security. They are mainly caused by the failures of the drivers to perceive the changes of the traffic lights or the unexpected conditions happening accidentally on the roads. In this paper, we devised a quantitative analysis for assessing driver's cognitive responses by investigating the neurobiological information underlying electroencephalographic (EEG) brain dynamics in traffic-light experiments in a virtual-reality (VR) dynamic driving environment. The VR technique allows subjects to interact directly with the moving virtual environment instead of monotonic auditory and visual stimuli, thereby provides interactive and realistic tasks without the risk of operating on an actual machine. Independent component analysis (ICA) is used to separate and extract noise-free ERP signals from the multi-channel EEG signals. A temporal filter is used to solve the time-alignment problem of ERP features and principle component analysis (PCA) is used to reduce feature dimensions. The dimension-reduced features are then input to a self-constructing neural fuzzy inference network (SONFIN) to recognize different brain potentials stimulated by red/green/yellow traffic events, the accuracy can be reached 87% in average eight subjects in this visual-stimuli ERP experiment. It demonstrates the feasibility of detecting and analyzing multiple streams of ERP signals that represent operators' cognitive states and responses to task events.     

A Novel Evolutionary-Based Cooperative Spectrum Sensing Mechanism for Cognitive Radio Networks

Cognitive radio (CR) is considered as a feasible intelligent technology for 4G wireless networks or self-organization networks and envisioned as a promising paradigm of exploiting intelligence for enhancing efficiency of underutilized spectrum bands. In CR, one of the main concerns is to reliably sense the presence of primary users, to attain protection against harmful interference caused by the potential spectrum access of secondary users (SUs). In this paper, evolutionary algorithms, namely, genetic algorithm (GA) and particle swarm optimization (PSO) are investigated. An imperialistic competitive algorithm (ICA) is proposed to minimize error detection at the common soft data fusion (SDF) center for structurally centralized cognitive radio network (CRN). By using these techniques, evolutionary operations are invoked to optimize the weighting coefficients applied on the sensing measurement components received from multiple cooperative SUs. The proposed method is compared with other evolutionary algorithms, as well as other conventional deterministic, such as maximal ratio combining- (MRC-), modified deflection coefficient- (MDC-), normal deflection coefficient- (NDC-) based SDF schemes and OR-rule HDF based. MATLAB simulations confirm the superiority of the ICA-based scheme over the PSO-, GA-based and other conventional schemes in terms of detection performance. In addition, the ICA-based scheme also shows promising convergence and time running performance as compared to other iterative-based schemes. This makes ICA an adequate solution to meet real-time requirements.     

一种基于局部密度的自适应眼电伪迹去除方法

脑电信号幅值微弱且信噪比低易受到多种伪迹影响.其中,眼电伪迹幅值高、随机性强,常使脑电信号产生明显畸变,对信号的后续分析将产生极大的影响.传统伪迹去除方法难以精确定位伪迹成分,导致过多有效信息丢失.针对上述问题,该文提出一种基于数据驱动的自适应伪迹定位和去除方法.该方法将局部密度引入独立成分分析(ICA)并通过聚类分析...     

基于盲源分离方法的工频干扰消除

生物医学信号检测和处理过程中,50Hz(或60Hz)工频干扰的消除一直非常重要的技术问题。本文提出了一种基于独立分量分析的工频干扰消除新方法。我们用该方法对模拟数据和实测脑电(EEG)信号中的工频干扰进行消除,并与自适应滤波工频干扰消除方法进行了比较。实验表明,本文所提出的方法在工频干扰消除上具有一定的优势。     

基于EOG的安全辅助驾驶系统算法设计与实现

为保证驾驶安全,提高车辆控制系统的智能化水平,实现“手不离盘”操作,设计并实现了一种基于眼电图(EOG)的安全辅助驾驶系统。该系统利用安装在驾驶员眼睛周围的生物电极采集其在观测抬头显示器(HUD, head up display)上提示符时所产生的扫视信号,生成多种车载设备控制命令;对原始多导联EOG信号进行端点检测后,使用了独立分量分析(ICA, independent component analysis)方法进行空域滤波后提取眼动信号特征参数,并结合支持向量机实现了上、左与右扫视动作的识别。实验室环境下对所提算法进行了测试,15位受试者在疲劳与非疲劳状态下的在线平均正确率达到了98.43%与96.0%。实验结果表明,基于ICA多类扫视信号识别算法的安全辅助驾驶系统在眼动信号分析中呈现出了良好的分类性能。     

基于车道信息融合的车辆行为识别

在车辆的自动驾驶和辅助驾驶中,实时分析车辆的运动状态具有重要的实际应用价值。为了实现对车辆行为的判断,提出一种基于车道信息融合的车辆行为识别算法。首先提出一种基于改进Robinson与LSD的模型,运用改进的Robinson算子获取最佳梯度幅值实现对车道的边缘提取,再通过LSD算法实现车道的检测。然后采用一种基于滑动窗口的三次样条插值法对车道进行拟合,最后根据车道参数信息分析车辆的运动状态,结合车辆的中心位置得到车辆的偏离信息。在BDD100K数据集的测试中,本文算法的车道检测准确率为95.61%,车辆行为识别准确率为93.04%,每秒传输帧数达到42.37。实验结果表明,本文算法在不同场景下可以有效地区分车辆的运动状态并给出车辆的偏离信息,具有更高的准确性和鲁棒性。     

基于光学多传感器场景信息的视觉辅助驾驶技术

为满足坦克、装甲车辆等军用车辆的闭舱、无窗驾驶需求,研制了一套新型辅助驾驶系统。系统将分布于车辆四周的多路光学传感器获取车辆近身场景,通过全景拼接算法得到车辆近身360°的全景鸟瞰视频,该视频显示于车载显示屏,用于车辆通过窄道、有障碍物等特殊路段,或倒车时,驾驶员观看。同时,在车辆通过常规路段时,上述视频可根据驾驶员头部扭转角度,裁选出符合人眼观察视角的车外场景视频,传输至驾驶员显示头盔上,供驾驶员观看。如遇特殊情况,车载显示屏会报警,驾驶员将回归车载显示屏的观看。其中,驾驶员头部位置确定方法采用了红外LED光源图像定位技术和MEMS惯性器件定位技术。在实验室,搭建模型小车,验证全景鸟瞰视频生成技术和头盔自由视点观察技术。此外,还使用真实车辆进行了跑车实验。实验结果表明,上述系统可满足闭舱无窗车辆在常规路况下行驶,速度可达40 km/h,同时可辅助车辆窄道行车、障碍物绕行和倒车等事项顺利进行。     

EEG-based prediction of driver's cognitive performance by deep convolutional neural network

We considered the prediction of driver's cognitive states related to driving performance using EEG signals. We proposed a novel channel-wise convolutional neural network (CCNN) whose architecture considers the unique characteristics of EEG data. We also discussed CCNN-R, a CCNN variation that uses Restricted Boltzmann Machine to replace the convolutional filter, and derived the detailed algorithm. To test the performance of CCNN and CCNN-R, we assembled a large EEG dataset from 3 studies of driver fatigue that includes samples from 37 subjects. Using this dataset, we investigated the new CCNN and CCNN-R on raw EEG data and also Independent Component Analysis (ICA) decomposition. We tested both within-subject and cross-subject predictions and the results showed CCNN and CCNN-R achieved robust and improved performance over conventional DNN and CNN as well as other non-DL algorithms.     

ICA-based multi-colour space texture classification system

Presented is a novel method which uses independent component analysis (ICA) for systematically partitioning and combining textural features extracted from different colour spaces, in a multiple classifier based system, for colour texture classification. Results obtained illustrate that the proposed ICA-based feature-partitioning and classifier combination system produces more accurate results compared to a system that combines classifiers applied to features extracted from individual colour spaces     

Discriminant Independent Component Analysis as a subspace representation

Subspace modeling plays an important role in face recognition. Independent Component Analysis （ICA）, a multivariable statistical analysis technique, can be seen as an extension of traditional Principal Com- ponent Analysis （PCA） technique, which addresses high order statistics as well as second order statistics. In this paper, a new scheme of subspace-based representation called Discriminant Independent Component Analysis （DICA） is proposed, which combines the strength＂ of unsupervised learning of ICA and supcrvised learning of Linear Discriminant Analysis （LDA）, and efficiently enhances the generalization ability of ICA-based representation method. Based on DICA subspace analysis, a set of optimal vectors called ＂discriminant independent faces＂ are learned from face samples. The effectiveness of our method is demonstrated by performance comparisons with some popular methods such as ICA, PCA, and PCA＋LDA. On the large scale database of IIS, significant improvements are observed when there are fewer training samples per person available.     

Trip-Based Optimal Power Management of Plug-in Hybrid Electric Vehicles

Hybrid electric vehicles (HEVs) have demonstrated the capability to improve fuel economy and emissions. The plug-in HEV (PHEV), utilizing more battery power, has become a more attractive upgrade of the HEV. The charge-depletion mode is more appropriate for the power management of PHEVs, i.e., the state of charge (SOC) is expected to drop to a low threshold when the vehicle reaches the trip destination. Trip information has so far been considered as future information for vehicle operation and is thus not available a priori. This situation can be changed by the recent advancement in intelligent transportation systems (ITSs) based on the use of on-board global positioning systems (GPSs), geographical information systems (GISs), and advanced traffic flow modeling techniques. In this paper, a new approach to optimal power management of PHEVs in the charge-depletion mode is proposed with driving cycle modeling based on the historic traffic information. A dynamic programming (DP) algorithm is applied to reinforce the charge-depletion control such that the SOC drops to a specific terminal value at the end of the driving cycle. The vehicle model was based on a hybrid electric sport utility vehicle (SUV). Only fuel consumption is considered for the current stage of the study. A simulation study was conducted for several standard driving cycles and two trip models using the proposed method, and the results showed significant improvement in fuel economy compared with a rule-based control and a depletion sustenance control for most cases. Furthermore, the results showed much better consistency in fuel economy compared with rule-based and depletion sustenance control.      

基于空间探索和认知图构建的生物启发式目标导向导航模型

为实现智能自主运行体面向目标的导航知识生成及运行控制,该文研究了一种基于空间探索和认知图构建的生物启发式目标导向(GO)导航模型,该模型由空间探索、认知图构建和GO导航控制3个部分组成。在空间探索中,将网格细胞(GCs)到位置细胞(PCs)模型和视觉位置细胞生成模型融合后生成的位置细胞表征当前状态,利用Q学习算法实现状态-动作的建立及更新,以此学习面向目标运行的导航知识;然后,在认知图构建中,利用重心估计原理对空间探索得到的知识进行处理,生成各位置细胞状态下面向目标的方向信息;最后,运行体在朝目标的运行中,根据得到的认知图实时控制运行方向,以此实现GO导航。仿真结果表明,该GO模型有效,运行体进行充分的空间探索可生成认知图,并以此实现GO导航,且在运行过程中能有效规避障碍物。     

基于图像独立特征分量的数字水印新方案

随着独立分量分析的方法被用来获取图像的独立特征分量，进行图像的分析，基于这种方法的数字水印技术也就随之出现。本文提出对原来的直接在图像的独立特征分量上加入水印的方法进行改进，对这些独立分量利用DCT进行去相关处理，在所得到的系数上嵌入水印，并且通过估计影响水印提取的各种因素，进一步提高鲁棒性。     

Autoregressive State Prediction Model Based on Hidden Markov and the Application

Considering the inaccuracies of the traditional Hidden Markov Model (HHM) in the dynamic processes that are close relatively related before and after characterization, an autoregressive state prediction model based on Hidden Markov with Autoregressive model and the coefficient of AR is proposed, which takes the coefficient of AR as the observations of the continuous HHM. Taking the recognition and prediction of heavy vehicle driving states as the research object, a two-layer HMM model is set up to describe the state of the whole steering process of the vehicle. The AR model is for the features extracting of the observations in a short period of time, and the coefficient of AR is extracted as the observed sequence of the lower HMM model library. The upper HMM is used to identify and predict the overall state of the vehicle during steering. The proposed model makes the state sequence with the highest probability on-line predicted in the observed sequence by the Viterbi algorithm, and calculates the state transition law to predict the state of the vehicle in a certain period of time in the future using the Markov prediction algorithm. Combining the double lane change and hook steering to train the parameters of the model, the online identification and prediction of heavy vehicle rollover states can be achieved. The results show that the proposed model can accurately identify the driving state of the vehicle with good real-time performance, and the good prediction on the trend of heavy vehicle driving conditions is verified.     

Multidimensional Blind Separation Using Higher-Order Statistics: Application to Non-Cooperative STBC Systems

Blindly separating the intercepted signals is a challenging problem in non-cooperative multiple input multiple output systems in association with space–time block code (STBC) where channel state information and coding matrix are unavailable. To our knowledge, there is no report on dealing with this problem in literature. In this paper, the STBC systems are represented with an independent component analysis (ICA) model by merging the channel and coding matrices as virtual channel matrix. Analysis shows that the source signals are of group-wise independence and the condition of mutual independence can not be satisfied for ordinary ICA algorithms when specific modulations are employed. A new multidimensional ICA algorithm is proposed to separate the intercepted signals in this case by jointly block-diagonalizing (JBD) the cumulant matrices. In this paper, JBD is achieved by a 2-step optimization algorithm and a contrast function is derived from the JBD criterion to remove the additional permutation ambiguity with explicit mathematical explanations. The convergence of the new method is guaranteed. Compared with the ICA-based channel estimation methods, simulations show that the new algorithm, which does not introduce additional ambiguities, achieves better performance with faster convergence in a non-cooperative scenario.