Search Results

Accurate perception of the driving environment and a highly accurate position of the vehicle are paramount to safe Autonomous Vehicle (AV) operation. AVs gather data about the environment using various sensors. For a robust perception and localization system, incoming data from multiple sensors is usually fused together using advanced computational algorithms, which historically requires a high-compute load. To reduce AV compute load and its negative effects on vehicle energy efficiency, we propose a new infrastructure information source (IIS) to provide environmental data to the AV. The new energy–efficient IIS, chip–enabled raised pavement markers are mounted along road lane lines and are able to communicate a unique identifier and their global navigation satellite system position to the AV. This new IIS is incorporated into an energy efficient sensor fusion strategy that combines its information with that from traditional sensor.

Heavy-duty commercial vehicles consume a significant amount of energy due to their large size and mass, directly leading to vehicle operators prioritizing energy efficiency to reduce operational costs and comply with environmental regulations. One tool that can be used for the evaluation of energy efficiency in heavy-duty vehicles is the evaluation of energy efficiency using vehicle modeling and simulation. Simulation provides a path for energy efficiency improvement by allowing rapid experimentation of different vehicle characteristics on fuel consumption without the need for costly physical prototyping. The research presented in this paper focuses on using real-world, sparsely sampled telematics data from a large fleet of heavy-duty vehicles to create high-fidelity models for simulation. Samples in the telematics dataset are collected sporadically, resulting in sparse data with an infrequent and irregular sampling rate.

Development of autonomous vehicle technology is expensive and perhaps more complicated than initially thought, as evidenced by the recent rollback of anticipated delivery dates from companies such as Tesla, Waymo, GM, and more. One of the most effective techniques to reduce research and development costs and speed up implementation is rigorous analysis through simulation. In this paper, we present multiple autonomous vehicle perception and control strategies that are rigorously investigated in the user friendly, free, and open-source simulation environment, CARLA. Overall, we successfully formulated potential solutions to the autonomous navigation problem and assessed their advantages and disadvantages in simulation at no cost. First, a lane finding method utilizing polynomial fitting and machine learning is proposed. Then, the waypoint navigation strategy is described, along with route planning. Object detection is then implemented using pre-trained convolutional neural networks.

The transportation sector contributes significantly to emissions and air pollution globally. Emission models of modern vehicles are important tools to estimate the impact of technologies or controls on vehicle emission reductions, but developing a simple and high-fidelity model is challenging due to the variety of vehicle classes, driving conditions, driver behaviors, and other physical and operational constraints. Recent literature indicates that neural network-based models may be able to address these concerns due to their high computation speed and high-accuracy of predicted emissions. In this study, we seek to expand upon this initial research by utilizing several deep neural networks (DNN) architectures such as a recurrent neural network (RNN) and a convolutional neural network (CNN). These DNN algorithms are developed specific to the vehicle-out emissions prediction application, and a comprehensive assessment of their performances is done.

While machine learning in autonomous vehicles development has increased significantly in the past few years, the use of reinforcement learning (RL) methods has only recently been applied. Convolutional Neural Networks (CNNs) became common for their powerful object detection and identification and even provided end-to-end control of an autonomous vehicle. However, one of the requirements of a CNN is a large amount of labeled data to inform and train the neural network. While data is becoming more accessible, these networks are still sensitive to the format and collection environment which makes the use of others’ data more difficult. In contrast, RL develops solutions in a simulation environment through trial and error without labeled data. Our research expands upon previous research in RL and Proximal Policy Optimization (PPO) and the application of these algorithms to 1/18th scale cars by expanding the application of this control strategy to a full-sized passenger vehicle.

Prediction of vehicle velocity is important since it can realize improvements in the fuel economy/energy efficiency, drivability, and safety. Velocity prediction has been addressed in many publications. Several references considered deterministic and stochastic approaches such as Markov chain, autoregressive models, and artificial neural networks. There are numerous new sensor and signal technologies like vehicle-to-vehicle and vehicle-to-infrastructure communication that can be used to obtain inclusive datasets. Using these inclusive datasets of sensors in deep neural networks, high accuracy velocity predictions can be achieved. This research builds upon previous findings that Long Short-Term Memory (LSTM) deep neural networks provide low error velocity prediction. We developed an LSTM deep neural network that uses different groups of datasets collected in Fort Collins, Colorado.

Signal processing incorporating Artificial Neural Networks (ANN) has been shown to be well suited for modeling engine-related performance indicators [ 1 , 2 , 3 ] that require multi-dimensional parametric calibration space. However, to obtain acceptable accuracy, traditional ANN implementation may require processing resources beyond the capability of current engine controllers. This paper explores the practicality of implementing an ANN-based algorithm performing real-time calculations of the volumetric efficiency (VE) for an engine with variable valve actuation (phasing and lift variation). This alternative approach was considered attractive since the additional degree of freedom introduced by variable lift would be cumbersome to add to the traditional multi-dimensional table-based representation of VE.

In racetrack conditions, brake systems are subjected to extreme energy loads and energy load distributions. This can lead to very high friction surface temperatures, especially on the brake corner that operates, for a given track, with the most available traction and the highest energy loading. Individual brake corners can be stressed to the point of extreme fade and lining wear, and the resultant degradation in brake corner performance can affect the performance of the entire brake system, causing significant changes in pedal feel, brake balance, and brake lining life. It is therefore important in high performance brake system design to ensure favorable operating conditions for the selected brake corner components under the full range of conditions that the intended vehicle application will place them under. To address this task in an early design stage, it is helpful to use brake system modeling tools to analyze system performance.

With increasingly stringent emissions regulations and concurrent requirements for enhanced engine thermal efficiency, a comprehensive characterization of the automotive gasoline fuel spray has become essential. The acquisition of accurate and repeatable spray data is even more critical when a combustion strategy such as gasoline direct injection is to be utilized. Without industry-wide standardization of testing procedures, large variablilities have been experienced in attempts to verify the claimed spray performance values for the Sauter mean diameter, Dv90, tip penetration and cone angle of many types of fuel sprays. A new SAE Recommended Practice document, J2715, has been developed by the SAE Gasoline Fuel Injection Standards Committee (GFISC) and is now available for the measurement and characterization of the fuel sprays from both gasoline direct injection and port fuel injection injectors.

This work discusses the development of SAE procedure J2747, “Hydraulic Pump Airborne Noise Bench Test”. This is a test procedure describing a standard method for measuring radiated sound power levels from hydraulic pumps of the type typically used in automotive power steering systems, though it can be extended for use with other types of pumps. This standard was developed by a committee of industry representatives from OEM's, suppliers and NVH testing firms familiar with NVH measurement requirements for automotive hydraulic pumps. Details of the test standard are discussed. The hardware configuration of the test bench and the configuration of the test article are described. Test conditions, data acquisition and post-processing specifics are also included. Contextual information regarding the reasoning and priorities applied by the development committee is provided to further explain the strengths, limitations and intended usage of the test procedure.

Tensile measurements and fracture surface analysis of low carbon heat-treated boron steel are reported. Tensile coupons were quasi-statically deformed to fracture in a miniature tensile testing stage with custom data acquisition software. Strain contours were computed via a digital image correlation method that allowed placement of a digital strain gage in the necking region. True stress-true strain data corresponding to the standard tensile testing method are presented for comparison with previous measurements. Fracture surfaces were examined using scanning electron microscopy and the deformation mechanisms were identified.

This paper describes the development of a mean value model for a turbocharged diesel engine. The objective is to develop a fast-running engine model with sufficient accuracy over a wide range of operating conditions for efficient evaluation of control algorithms and control strategies. The mean value engine model was derived from a detailed 1D engine model, using the Design of Experiments (DOE) and hybrid Radial Basis Functions (RBF) to approximate the simulation results of the detailed model for cylinder quantities (e.g., the engine volumetric efficiency, the indicated efficiency, and the energy fraction of the exhaust gas). Furthermore, the intake and exhaust systems (especially intake and exhaust manifolds) were completely simplified by lumping flow components together. In addition, to compare with hybrid RBF, neural networks were also used to approximate the simulation results of the detailed engine model.

For automotive applications at elevated temperatures, the need for sufficient creep resistance of Mg alloys is often associated with retaining appropriate percentages of initial clamp loads in bolt joints. This engineering property is often referred to as bolt-load retention (BLR); BLR testing is a practical method to quantify the bolt load with time for engineering purposes. Therefore, standard BLR test procedures for automotive applications are desired. This report summarizes the effort in the Structural Cast Magnesium Development (SCMD) project under the United States Automotive Materials Partnership (USAMP), to provide a technical basis for recommending a general-purpose and a design-purpose BLR test procedures for BLR testing of Mg alloys for automotive applications. The summary includes results of factors influencing BLR and related test techniques from open literature, automotive industry and research carried out in this laboratory project.

This paper is targeted to engineers who are involved in predicting fatigue life using either the strain-life approach or the stress-life approach. However, more emphasis is given to the strain-life approach, which is commonly used for fatigue life analysis in the ground vehicle industry. It attempts to discuss, modify and extend approaches in fatigue analysis, so they are best suited for structural durability engineers. Fatigue analysis requires the use of material fatigue properties, stress or strain results obtained from finite element analyses or measurements, and load data obtained from multi-body dynamic analysis or road load data acquisition. This paper examines the effects of these variables in predicting fatigue life. Various mean stress corrections, along with their advantages and disadvantages are discussed. Different stress/strain combinations such as signed von Mises, and signed Tresca are examined. Also, advanced methods such as Fatemi-Socie and Bannantine are discussed.