Search Results

Controller Area Network (CAN), the de facto standard for in-vehicle networks, has insufficient security features and thus is inherently vulnerable to various attacks. To protect CAN bus from attacks, intrusion detection systems (IDSs) based on advanced deep learning methods, such as Convolutional Neural Network (CNN) and Recurrent Neural Network (RNN), have been proposed to detect intrusions. However, those models generally introduce high latency, require considerable memory space, and often result in high energy consumption. To accelerate intrusion detection and also reduce memory requests, we exploit the use of Binarized Neural Network (BNN) and hardware-based acceleration for intrusion detection in in-vehicle networks. As BNN uses binary values for activations and weights rather than full precision values, it usually results in faster computation, smaller memory cost, and lower energy consumption than full precision models.

There are many methods developed over the past decade to solve the problem of energy management control for hybrid electric vehicles. A novel method is introduced in this paper to address the same problem which reduces the problem to a set of physical equations and maps. In simple terms, this method directly calculates the actual cost or savings in fuel energy from the generation or usage of electric energy. It also calculates the local optimum electric power that yields higher electric fuel savings (EFS) or lower electric fuel cost (EFC) in the fuel energy that is spent for driving the vehicle (which in general does not take the system to the lowest engine Brake Specific Fuel Consumption (BSFC)). Based on this approach, a control algorithm is developed which attempts to approach the global optimum over a drive cycle.

The design of a demonstration vehicle is presented where improvements to the electrical and air induction systems are made which provide increased performance with improved fuel economy. This is made possible by a 48 V architecture which enables the deployment of new components, specifically a belted motor generator unit (MGU) and electrically-driven compressor (eBooster®). The synergy between these components enables energy efficient means to collect regenerated energy and provide added torque, faster engine response, and extended engine off operation among a list of added features. Control features and strategy are highlighted along with simulation and vehicle test data. Resultant performance and fuel economy benefits are reviewed which support the contention of 48 V being a cost effective architecture to enable CO2 reduction relative to a higher voltage hybrid.

The Controls and Software Engineering Team at BorgWarner Drivetrain Systems has successfully employed model-based software development for the past several years. Their drivetrain system control software, developed using MATLAB/Simulink/Stateflow, and autocoded using TargetLink, is on the road in many passenger vehicle applications. Using these tools, BorgWarner has realized the widely recognized benefits of model-based design; such as increased speed to market, improved quality, and reduced complexity. Validating algorithms early through simulation and rapid prototyping, then translating them to production software through automatic code generation has proven very successful for BorgWarner. When starting with model-based design, the BorgWarner team focused on developing the core application control algorithms in the modeling environment. Lower-level software such as I/O drivers, the task scheduler, and communication logic was still hand-coded.

In military vehicles reliability can sometimes be a more important issue than cost. With that in perspective, this paper discusses the load availability and reliability issues in automotive multiplexed wiring harness systems, which are potentially useful in the military, and compares the same with a regular non-multiplexed system. For that purpose, a figure of merit or metric is introduced, and the load availability is described in terms of this metric, which depends on the architecture chosen.