A Recognition Method of Aggressive Driving Behavior Based on Ensemble Learning

Aggressive driving behavior (ADB) is one of the main causes of traffic accidents. The accurate recognition of ADB is the premise to timely and effectively conduct warning or intervention to the driver. There are some disadvantages, such as high miss rate and low accuracy, in the previous data-driven recognition methods of ADB, which are caused by the problems such as the improper processing of the dataset with imbalanced class distribution and one single classifier utilized. Aiming to deal with these disadvantages, an ensemble learning-based recognition method of ADB is proposed in this paper. First, the majority class in the dataset is grouped employing the self-organizing map (SOM) and then are combined with the minority class to construct multiple class balance datasets. Second, three deep learning methods, including convolutional neural networks (CNN), long short-term memory (LSTM), and gated recurrent unit (GRU), are employed to build the base classifiers for the class balance datasets. Finally, the ensemble classifiers are combined by the base classifiers according to 10 different rules, and then trained and verified using a multi-source naturalistic driving dataset acquired by the integrated experiment vehicle. The results suggest that in terms of the recognition of ADB, the ensemble learning method proposed in this research achieves better performance in accuracy, recall, and F1-score than the aforementioned typical deep learning methods. Among the ensemble classifiers, the one based on the LSTM and the Product Rule has the optimal performance, and the other one based on the LSTM and the Sum Rule has the suboptimal performance.

Driver’s Personality and Behavior for Boosting Automobile Security and Sensing Health Problems Through Fuzzy Signal Detection Case Study: Mexico City

Automobile security became an essential theme over the last years, and some automakers invested much money for collision avoidance systems, but personalization of their driving systems based on the user’s behavior was not explored in detail. Furthermore, efficiency gains could be had with tailored systems. In Mexico, 80% of automobile accidents are caused by human beings; the remaining 20% are related to other issues such as mechanical problems. Thus, 80% represents a significant opportunity to improve safety and explore driving efficiency gains. Moreover, when driving aggressively, it could be connected with mental health as a post-traumatic stress disorder. This paper proposes a Tailored Collision Mitigation Braking System, which evaluates the driver’s personality driving treats through signal detection theory to create a cognitive map that understands the driving personality of the driver. In this way, aggressive driving can be detected; the system is then trained to recognize the personality trait of the driver and select the appropriate stimuli to achieve the optimal driving output. As a result, when aggressive driving is detected continuously, an automatic alert could be sent to the health specialists regarding particular risky behavior linked with mental problems or drug consumption. Thus, the driving profile test could also be used as a detector for health problems.

Model-Based Range Prediction for Electric Cars and Trucks under Real-World Conditions

The further development of electric mobility requires major scientific efforts to obtain reliable data for vehicle and drive development. Practical experience has repeatedly shown that vehicle data sheets do not contain realistic consumption and range figures. Since the fear of low range is a significant obstacle to the acceptance of electric mobility, a reliable database can provide developers with additional insights and create confidence among vehicle users. This study presents a detailed, yet easy-to-implement and modular physical model for both passenger and commercial battery electric vehicles. The model takes consumption-relevant parameters, such as seasonal influences, terrain character, and driving behavior, into account. Without any a posteriori parameter adjustments, an excellent agreement with known field data and other experimental observations is achieved. This validation conveys much credibility to model predictions regarding the real-world impact on energy consumption and cruising range in standardized driving cycles. Some of the conclusions, almost impossible to obtain experimentally, are that winter conditions and a hilly terrain each reduce the range by 7–9%, and aggressive driving reduces the range by up to 20%. The quantitative results also reveal the important contributions of recuperation and rolling resistance towards the overall energy budget.

Development of a Scaled Vehicle Model for Dynamics Testing

The interest in using scaled models for dynamics testing of prototype vehicles is growing due to the high demand for autonomous driving. In the early design phases, vehicle testing is done using computer simulations. Even though computer simulations are proven to be extremely helpful in designing prototypes, simulation models need to be validated using realworld testing. There are high costs involved in vehicle testing and it’s dangerous to conduct aggressive driving manoeuvres with real drivers. As a solution, researchers have used scaled models. To validate the computer simulations, researchers matched the scaled model test data with full-size vehicle prototypes considering the dynamic similitude. However, previous work was limited to the analysis of the steady-state behaviour of vehicles. To accurately predict the behaviour, the transientstate response must be tested as well. Therefore, this paper outlines the precursory work of a scaled model with the ability to test both states during vehicle manoeuvres. This paper is structured as follows. Section II presents related work. Section III elaborates on the mathematical modelling and present the results of the computer simulations. Section IV presents the scaled model which will be developed. Section V concludes the findings, and present the future work of research.

Understanding Aggressive Driving: The Role of Personality and Individual Differences

The goal of the present study was to examine the role of personality and individual differences on aggressive driving. It was hypothesized that personality and individual differences would be significantly related to aggressive driving behavior. A sample of n = 252 participants from a southeastern university and surrounding community were required to complete a series of driving questionnaires; the ADBQ, DBQ, and CFQ-D; and a series of personality questionnaires; the IPIP-NEO-PIR and BFI. Our results indicated that personality factors and individual differences significantly predicted aggressive driving outcomes. These results provided a preliminary personality based characteristic profile of the aggressive driver. These results also support the use of trait anger and trait cooperation independently from the subscales they are derived from (Neuroticism and Agreeableness) to predict aggressive driving behaviors. Theoretical and practical implications are discussed.