An Integrated LSTM-Rule-Based Fusion Method for the Localization of Intelligent Vehicles in a Complex Environment.

To improve the accuracy and robustness of autonomous vehicle localization in a complex environment, this paper proposes a multi-source fusion localization method that integrates GPS, laser SLAM, and an odometer model. Firstly, fuzzy rules are constructed to accurately analyze the in-vehicle localization deviation and confidence factor to improve the initial fusion localization accuracy. Then, an odometer model for obtaining the projected localization trajectory is constructed. Considering the high accuracy of the odometer's projected trajectory within a short distance, we used the shape of the projected localization trajectory to inhibit the initial fusion localization noise and used trajectory matching to obtain an accurate localization. Finally, the Dual-LSTM network is constructed to predict the localization and build an electronic fence to guarantee the safety of the vehicle while also guaranteeing the updating of short-distance localization information of the vehicle when the above-mentioned fusion localization is unreliable. Under the limited arithmetic condition of the vehicle platform, accurate and reliable localization is realized in a complex environment. The proposed method was verified by long-time operation on the real vehicle platform, and compared with the EKF fusion localization method, the average root mean square error of localization was reduced by 66%, reaching centimeter-level localization accuracy.

Emergency Braking Evoked Brain Activities during Distracted Driving.

Electroencephalogram (EEG) was used to analyze the mechanisms and differences in brain neural activity of drivers in visual, auditory, and cognitive distracted vs. normal driving emergency braking conditions. A pedestrian intrusion emergency braking stimulus module and three distraction subtasks were designed in a simulated experiment, and 30 subjects participated in the study. The common activated brain regions during emergency braking in different distracted driving states included the inferior temporal gyrus, associated with visual information processing and attention; the left dorsolateral superior frontal gyrus, related to cognitive decision-making; and the postcentral gyrus, supplementary motor area, and paracentral lobule associated with motor control and coordination. When performing emergency braking under different driving distraction states, the brain regions were activated in accordance with the need to process the specific distraction task. Furthermore, the extent and degree of activation of cognitive function-related prefrontal regions increased accordingly with the increasing task complexity. All distractions caused a lag in emergency braking reaction time, with 107.22, 67.15, and 126.38 ms for visual, auditory, and cognitive distractions, respectively. Auditory distraction had the least effect and cognitive distraction the greatest effect on the lag.

An experimental investigation of particle and NOx emissions for a non-road diesel engine equipped with an integrated DOC + CDPF + SCR aftertreatment system during different operations.

An integrated aftertreatment system consisting of diesel oxidation catalyst (DOC), catalytic diesel particulate filter (CDPF), and selective catalytic reduction (SCR) is an effective way of reducing both NOx and particulate matter (PM). In this paper, the effect of DOC + CDPF + SCR on NOX and particle emissions is investigated during different operations to assess applicability of this aftertreatment for meeting more rigorous non-road emissions standard. Meanwhile non-negligible issue about regeneration emission is studied. The results show that the DOC + CDPF have no significant effect on NOx but increase the NO2/NOx ratio which is correlated with load. SCR is the main NOx reduction device with average efficiency of 86.5% for steady-state operations. Meanwhile, NH3 slip is lower than 16 ppm. During cold and hot non-road transient cycles (NRTC cycles), average NOx efficiencies are 56.7% and 57.8%, respectively, along with NH3 slip below 10 ppm. DOC + CDPF + SCR maintain filtration efficiency over 97% and 99% for PM and particle number (PN) for either steady-state operation or NRTC cycle, but particle size distributions change. Compared with the original emissions, NOx, PM, and PN emission factors are all below non-road China-IV limit after equipping with DOC + CDPF + SCR. However, during regeneration the aftertreatment does not maintain a high filtration performance but becomes particle generator. The penetration of nuclear particles and decomposition of agglomerated particles lead to high CDPF-out PN of 1.5 × 107 #/cm3-3.5 × 107 #/cm3. During regeneration, accumulated NOx is negligible, but the PM is 121.6 and 44.5 times higher than cold and hot NRTC cycles, respectively. In summary, DOC + CDPF + SCR is excellent way to improve non-road emissions but low SCR efficiency at low-temperature and high accumulated PM during regeneration process should be further addressed.

Advanced Sensing and Control for Connected and Automated Vehicles.

In recent years, connected and automated vehicles (CAV) have been a transformative technology that is expected to reduce emissions and change and improve the safety and efficiency of the mobilities [...].

A Laser-Based Multipass Absorption Sensor for Sub-ppm Detection of Methane, Acetylene and Ammonia.

A compact, sensitive laser-based absorption sensor for multispecies monitoring of methane (CH4), acetylene (C2H2) and ammonia (NH3) was developed using a compact multipass gas cell. The gas cell is 8.8 cm long and has an effective optical path length of 3.0 m with a sampling volume of 75 mL. The sensor is composed of three fiber-coupled distributed feedback lasers operating near 1512 nm, 1532 nm and 1654 nm, an InGaAs photodetector and a custom-designed software for data acquisition, signal processing and display. The lasers were scanned over the target absorption features at 1 Hz. First-harmonic-normalized wavelength modulation spectroscopy (f = 3 kHz) with the second harmonic detection (WMS-2f/1f) is employed to eliminate the unwanted power fluctuations of the transmitted laser caused by aerosol/particles scattering, absorption and beam-steering. The multispecies sensor has excellent linear responses (R2 > 0.997) within the gas concentration range of 1-1000 ppm and shows a detection limit of 0.32 ppm for CH4, 0.16 ppm for C2H2 and 0.23 ppm for NH3 at 1 s response time. The Allan-Werle deviation analysis verifies the long-term stability of the sensor, indicating a minimal detection limit of 20-34 ppb were achieved after 60-148 s integration time. Flow test of the portable multispecies sensor is also demonstrated in this work.

Simultaneous estimation of ammonia injection rate and state of diesel urea-SCR system based on high gain observer.

Selective catalytic reduction (SCR) controller achieves high NOx conversion and low NH3 slip primarily by means of controlling the rate and timing of injection of urea solution. No any low-cost commercial sensor can directly measure the actual ammonia (NH3) injection rate currently; thus, it is necessary to propose a numerical method to estimate the actual NH3 injection rate, which helps stabilization of emission control performances of a diesel SCR system during its service life cycle. A three-state nonlinear model was established and a method for designing high-gain observer (HGO) was studied so that those state variables such as NH3 injection rate, NOx concentrations, NH3 concentrations and NH3 coverage ratio can be estimated simultaneously. Such design method only requires bounded unknown input dynamics rather than any assumption of the fact how the unknown input changes. The simulation verification was performed by the ETC test cycles and qualitative and quantitative analysis of estimation errors were carried out. Compared with the extended Kalman filter (EKF) observer, the proposed high-gain observer has a higher estimation accuracy and better target traceability.

An Evaluation of Executive Control Function and Its Relationship with Driving Performance.

The driver's attentional state is a significant human factor in traffic safety. The executive control process is a crucial sub-function of attention. To explore the relationship between the driver's driving performance and executive control function, a total of 35 healthy subjects were invited to take part in a simulated driving experiment and a task-cuing experiment. The subjects were divided into three groups according to their driving performance (aberrant driving behaviors, including lapses and errors) by the clustering method. Then the performance efficiency and electroencephalogram (EEG) data acquired in the task-cuing experiment were compared among the three groups. The effect of group, task transition types and cue-stimulus intervals (CSIs) were statistically analyzed by using the repeated measures analysis of variance (ANOVA) and the post hoc simple effect analysis. The subjects with lower driving error rates had better executive control efficiency as indicated by the reaction time (RT) and error rate in the task-cuing experiment, which was related with their better capability to allocate the available attentional resources, to express the external stimuli and to process the information in the nervous system, especially the fronto-parietal network. The activation degree of the frontal area fluctuated, and of the parietal area gradually increased along with the increase of CSI, which implied the role of the frontal area in task setting reconstruction and working memory maintaining, and of the parietal area in stimulus-Response (S-R) mapping expression. This research presented evidence of the close relationship between executive control functions and driving performance.

Occlusion-Free Road Segmentation Leveraging Semantics for Autonomous Vehicles.

The deep convolutional neural network has led the trend of vision-based road detection, however, obtaining a full road area despite the occlusion from monocular vision remains challenging due to the dynamic scenes in autonomous driving. Inferring the occluded road area requires a comprehensive understanding of the geometry and the semantics of the visible scene. To this end, we create a small but effective dataset based on the KITTI dataset named KITTI-OFRS (KITTI-occlusion-free road segmentation) dataset and propose a lightweight and efficient, fully convolutional neural network called OFRSNet (occlusion-free road segmentation network) that learns to predict occluded portions of the road in the semantic domain by looking around foreground objects and visible road layout. In particular, the global context module is used to build up the down-sampling and joint context up-sampling block in our network, which promotes the performance of the network. Moreover, a spatially-weighted cross-entropy loss is designed to significantly increases the accuracy of this task. Extensive experiments on different datasets verify the effectiveness of the proposed approach, and comparisons with current excellent methods show that the proposed method outperforms the baseline models by obtaining a better trade-off between accuracy and runtime, which makes our approach is able to be applied to autonomous vehicles in real-time.

Correlation Among Behavior, Personality, and Electroencephalography Revealed by a Simulated Driving Experiment.

Drivers play the most important role in the human-vehicle-environment system and driving behaviors are significantly influenced by the cognitive state of the driver and his/her personality. In this paper, we aimed to explore the correlation among driving behaviors, personality and electroencephalography (EEG) using a simulated driving experiment. A total of 36 healthy subjects participated in the study. The 64-channel EEG data and the driving data, including the real-time position of the vehicle, the rotation angle of the steering wheel and the speed were acquired simultaneously during driving. The Cattell 16 Personality Factor Questionnaire (16PF) was utilized to evaluate the personalities of subjects. Through hierarchical clustering of the 16PF personality traits, the subjects were divided into four groups, i.e., the Inapprehension group, Insensitivity group, Apprehension group and the Unreasoning group, named after their representative personality trait. Their driving performance and turning behaviors were compared and EEG preprocessing, source reconstruction and the comparisons among the four groups were performed using Statistical Parameter Mapping (SPM). The turning process of the subjects can be formulated into two steps, rotating the steering wheel toward the turning direction and entering the turn, and then rotating the steering wheel back and leaving the turn. The bilateral frontal gyrus was found to be activated when turning left and right, which might be associated with its function in attention, decision-making and executive control functions in visual-spatial and visual-motor processes. The Unreasoning group had the worst driving performance with highest rates of car collision and the most intensive driving action, which was related to a higher load of visual spatial attention and decision making, when the occipital and superior frontal areas played a very important role. Apprehension (O) and Tension (Q4) had a positive correlation, and Reasoning (B) had a negative correlation with dangerous driving behaviors. Our results demonstrated the close correlation among driving behaviors, personality and EEG and may be taken as a reference for the prediction and precaution of dangerous driving behaviors in people with specific personality traits.

Behavior Evaluation Based on Electroencephalograph and Personality in a Simulated Driving Experiment.

Assessments and predictions of driving behavior are very important to improve traffic safety. We hypothesized that there were some patterns of driving behaviors, and these patterns had some correlation with cognitive states and personalities. To test this hypothesis, an evaluation of driving status, based on electroencephalography (EEG) and steering behavior in a simulated driving experiment, was designed and performed. Unity 3D was utilized to design the simulated driving scene. A photoelectric encoder fixed on the steering wheel and the corresponding data collection, transmission, and storage device was developed by Arduino, to acquire the rotation direction, angle, angular velocity, and angular acceleration of the steering wheel. Biopac MP 150 was utilized to collect the EEG data simultaneously during driving. A total of 23 subjects (mean age 23.6 ± 1.3 years, driving years: 2.4 ± 1.6 years, 21 males and two females) participated in this study. The Fuzzy C-means algorithm (FCMA) was utilized to extract patterns of driving behavior and the cognitive state within the window width of 20 s. The behaviors were divided into five kinds, i.e., negative, normal, alert, stress, and violent behavior, respectively, based on the standard deviation of steering wheel data. The cognitive states were divided into four kinds, i.e., negative, calm, alert, and tension, respectively, based on the EEG data. The correlation of these data, together with the personality traits evaluated using Cattell 16 Personality Factor Questionnaire (16PF) were analyzed using multiclass logistic regression. Results indicated the significance of the cognitive state and seven personality traits [apprehension (O), rule consciousness (G), reasoning (B), emotional stability (C), liveliness (F), vigilance (L), and perfectionism (Q3)] in predicting driving behaviors, and the prediction accuracy was 80.2%. The negative and alert cognitive states were highly correlated with dangerous driving, including negative and violent behaviors. Personality traits complicate the relationship with driving behaviors, which may vary across different types of subjects and traffic accidents.

Driving Style Recognition Based on Electroencephalography Data From a Simulated Driving Experiment.

Driving style is a very important indicator and a crucial measurement of a driver's performance and ability to drive in a safe and protective manner. A dangerous driving style would possibly result in dangerous behaviors. If the driving styles can be recognized by some appropriate classification methods, much attention could be paid to the drivers with dangerous driving styles. The driving style recognition module can be integrated into the advanced driving assistance system (ADAS), which integrates different modules to improve driving automation, safety and comfort, and then the driving safety could be enhanced by pre-warning the drivers or adjusting the vehicle's controlling parameters when the dangerous driving style is detected. In most previous studies, driver's questionnaire data and vehicle's objective driving data were utilized to recognize driving styles. And promising results were obtained. However, these methods were indirect or subjective in driving style evaluation. In this paper a method based on objective driving data and electroencephalography (EEG) data was presented to classify driving styles. A simulated driving system was constructed and the EEG data and the objective driving data were collected synchronously during the simulated driving. The driving style of each participant was classified by clustering the driving data via K-means. Then the EEG data was denoised and the amplitude and the Power Spectral Density (PSD) of four frequency bands were extracted as the EEG features by Fast Fourier transform and Welch. Finally, the EEG features, combined with the classification results of the driving data were used to train a Support Vector Machine (SVM) model and a leave-one-subject-out cross validation was utilized to evaluate the performance. The SVM classification accuracy was about 80.0%. Conservative drivers showed higher PSDs in the parietal and occipital areas in the alpha and beta bands, aggressive drivers showed higher PSD in the temporal area in the delta and theta bands. These results imply that different driving styles were related with different driving strategies and mental states and suggest the feasibility of driving style recognition from EEG patterns.

Formation of bicyclic polycyclic aromatic hydrocarbons (PAHs) from the reaction of a phenyl radical with cis-3-penten-1-yne.

The formation of polycyclic aromatic hydrocarbons (PAHs) on the C11H11 potential energy surface involved in the reactions of a phenyl radical (C6H5) with cis-3-penten-1-yne (cis-C1H[triple bond, length as m-dash]C2-C3H[double bond, length as m-dash]C4H-C5H3, referred to as C5H6) and its three radicals (CH[triple bond, length as m-dash]C-C[double bond, length as m-dash]CH-CH3, CH[triple bond, length as m-dash]C-CH[double bond, length as m-dash]C-CH3, and cis-CH[triple bond, length as m-dash]C-CH[double bond, length as m-dash]CH-CH2, referred to as the C3-, C4-, and C5-radicals with the same chemical components, C5H5) assisted by H atoms is investigated by performing combined density functional theory (DFT) and ab initio calculations. Five potential pathways for the formation of PAHs have been explored in detail: Pathways I-II correspond to the reaction of C6H5 with C5H6 at the C1 and C2 position, and Pathways III-V involve the reaction of C6H5 with the C3-, C4-, and C5-radicals with the assistance of H atoms. The initial association of C6H5 with C5H6 or C5H5 is found to be highly exothermic with only minor barriers (1.4-7.1 kcal mol-1), which provides a large driving force for the formation of PAHs. The hydrogen atom is beneficial for the ring enlargement and ring formation processes. The present calculations predict 9 potential PAHs, six (CS6, CS10, CS13, CS26, CS28 and CS29) of which are indicated to be energetically more favorable along Pathways I, III, IV and V at low temperature. The calculated barriers for the formation of these PAHs are around 19.2-38.0 kcal mol-1. All PAHs products could be formed at flame temperature, for the medium barriers are easily overcome in various flame conditions. The theoretical results supplement the PAH formation pathway and provide help to understand PAH growth mechanism.