Search Results

Evaluating real-world hazards associated with perception subsystems is critical in enhancing the performance of autonomous vehicles. The reliability of autonomous vehicles perception subsystems are paramount for safe and efficient operation. While current studies employ different metrics to evaluate perception subsystem failures in autonomous vehicles, there still exists a gap in the development and emphasis on engineering requirements. To address this gap, this study proposes the establishment of engineering requirements that specifically target real-world hazards and resilience factors important to AV operation, using High-Definition Maps, Global Navigation Satellite System, and weather sensors. The findings include the need for engineering requirements to establish clear criteria for a high-definition maps functionality in the presence of erroneous perception subsystem inputs which enhances the overall safety and reliability of the autonomous vehicles.

Goods movement and port related activities are a significant source of emissions in many large urban areas. Electrification of diesel cargo handling equipment is one method of reducing community exposure to these emissions, that also provides the potential for reducing greenhouse gas emissions. This study evaluated the performance of several pieces of zero emission cargo transfer equipment for a demonstration conducted at two terminal locations at the Port of Long Beach (POLB). This included the data logging of three battery-electric top handlers and one battery-electric yard tractor, as well as two baseline diesel top handlers and one diesel yard tractor. The battery-electric equipment typically operated about 5 hours per day, while using between 34 to 50% of the battery pack state of charge (SOC). In general, the battery-electric equipment was able to provide comparable hours of operation to the diesel equipment over a typical 8-hour shift.

This paper presents an energy-optimal deceleration planning system (EDPS) to maximize regenerative energy for electrified vehicles on deceleration events perceived by map and navigation information, machine vision and connected communication. The optimization range for EDPS is restricted within an upcoming deceleration event rather than the entire routes while in real time considering preceding vehicles. A practical force balance relationship based on an electrified powertrain is explicitly utilized for building a cost function of the associated optimal control problem. The optimal inputs are parameterized on each computation node from a set of available deceleration profiles resulting from a deceleration time model which are configured by real-world test drivings.

Eco-Approach and Departure (EAD) has been considered as a promising eco-driving strategy for vehicles traveling in an urban environment, where information such as signal phase and timing (SPaT) and geometric intersection description is well utilized to guide vehicles passing through intersections in the most energy-efficient manner. Previous studies formulated the optimal trajectory planning problem as finding the shortest path on a graphical model. While this method is effective in terms of energy saving, its computation efficiency can be further enhanced by adopting machine learning techniques. In this paper, we propose an innovative deep learning-based queue-aware eco-approach and departure (DLQ-EAD) system for a plug-in hybrid electric bus (PHEB), which is able to provide an online optimal trajectory for the vehicle considering both the downstream traffic condition (i.e. traffic lights, queues) and the vehicle powertrain efficiency.

Drivers continually require steering performance improvement, particularly in the area of feedback from the road. In this study, we develop a new electrically-assisted steering logic by 1) analyzing existing steering systems to determine key factors, 2) modeling an ideal steering system from which to obtain a desirable driver torque, 3) developing a rack force observer to faithfully represent road information and 4) building a feedback compensator to track the tuned torque. In general, the estimator uses the driver torque, assist torque and other steering system signals. However, the friction of the steering system is difficult to estimate accurately. At high speed, where steering feeling is very important, greater friction results in increased error. In order to solve this problem, we design two estimators generated from a vehicle model and a steering system model. The observer that uses two estimators can reflect various operating conditions by using the strengths of each method.

This paper compares the material consumption and fire patterns which developed on four nearly identical compact sedans when each was burned for exactly the same amount of time, but with different wind speed and direction during the burns. This paper will also compare the effects of environmental exposure to the fire patterns on the vehicles. The burn demonstrations were completed at an outdoor facility in southeast Michigan on four late model compact sedans. The wind direction was controlled by placing the subject vehicle with either the front facing into the wind, or rear facing into the wind. Two of the burns were conducted when the average observed wind speed was 5-6kph and two of the burns were conducted at an average observed wind speed of 19kph.

Wire shorts on an in-vehicle controller area network (CAN) impact the communication between electrical control units (ECUs), and negatively affects the vehicle control. The fault, especially the intermittent fault, is difficult to locate. In this paper, an equivalent circuit model for in-vehicle CAN bus is developed under the wire short fault scenario. The bus resistance is estimated and a resistance-distance mapping approach is proposed to locate the fault. The proposed approach is implemented in an Arduino-based embedded system and validated on a vehicle frame. The experimental results are promising. The approach presented in this paper may reduce trouble shooting time for CAN wire short faults and may enable early detection before the customer is inconvenienced.

Aerodynamic vehicle design improvements require flow simulation driven iterative shape changes. The 3-D flow field simulations (CFD analysis) are not explicitly descriptive in providing the direction for aerodynamic shape changes (reducing drag force or increasing the down-force). In recent times, aerodynamic shape optimization using the adjoint method has been gaining more attention in the automotive industry. The traditional DOE (Design of Experiment) optimization method based on the shape parameters requires a large number of CFD flow simulations for obtaining design sensitivities of these shape parameters. The large number of CFD flow simulations can be significantly reduced if the adjoint method is applied. The main purpose of the present study is to demonstrate and validate the adjoint method for vehicle aerodynamic shape improvements.

Optimizing/maximizing regen braking in a hybrid electric vehicle (HEV) is one of the key features for increasing fuel economy. However, it is known [1] that maximizing regen braking by braking the rear axle on a low friction surface results in compromising vehicle stability even in a vehicle which is equipped with an ESP (Enhanced Stability Program). In this paper, we develop a strategy to maximize regen braking without compromising vehicle stability. A yaw rate stability control system is designed for a hybrid electric vehicle with electric rear axle drive (ERAD) and a “hang on” center coupling device which can couple the front and rear axles for AWD capabilities. Nonlinear models of the ERAD drivetrain and vehicle are presented using bond graphs while a high fidelity model of the center coupling device is used for simulation.

Development of the next generation internal combustion engines and automatic transmissions for automotive applications is a mandatory powertrain engineering activity required now and in the coming years to meet forthcoming global emissions regulations. This paper details a preliminary investigation into possible synergies for fuel consumption reduction considering emerging automotive technologies integrated into the next generation combustion engine and automatic transmission architectures. A range of hypothetical gasoline engines were created and paired with a generalized set of step gear automatic transmissions designed to meet the performance requirements of high volume longitudinal full size truck application. These designs were then run through a design of experiments orthogonal array for prediction of fuel consumption on the WLTP test schedule and stand still acceleration to 100 kph.

Laser welding of dissimilar metals such as Aluminum and Copper, which is required for Li-ion battery joining, is challenging due to the inevitable formation of the brittle and high electrical-resistant intermetallic compounds. Recent research has shown that by using a novel technology, called laser braze-welding, the Al-Cu intermetallics can be minimized to achieve superior mechanical and electrical joint performance. This paper investigates the robustness of the laser braze-welding process. Three product and process categories, i.e. choice of materials, joint configurations, and process conditions, are studied. It is found that in-process effects such as sample cleanness and shielding gas fluctuations have a minor influence on the process robustness. Furthermore, many pre-process effects, e.g. design changes such as multiple layers or anodized base material can be successfully welded by process adaption.

The measurement of SO2 levels in vehicle exhaust can provide important information in understanding the relative contribution of sulfur and sulfate from fuel vs. oil source to PM. For this study, a differential optical absorption spectrometer (DOAS) that can measure SO2 down to 20 ppbV in real-time was built and evaluated. The DOAS consisted of an extractive sampling train, a cylindrical sampling cell with a single-path design to minimize cell volume, a spectrometer, and a deuterium lamp light source with a UVC range of ∼200-230 nanometer (nm). Laboratory tests showed detection limits were approximately in the range of 12 to 15 ppbV and showed good linearity over SO2 concentration ranges of 20 to 953 ppbV. Interference tests showed some interference by NO and by NH3, at levels of 300 ppmV and 16.6 ppmV, respectively.

Partially automated driving involves the relinquishment of longitudinal and/or latitudinal control to the vehicle. Partially automated systems, however, are fallible and require driver oversight to avoid all road hazards. Researchers have expressed concern that automation promotes extended eyes-off-road (EOR) behavior that may lead to a loss of situational awareness (SA), degrading a driver’s ability to detect hazards and make necessary overrides. A potential countermeasure to visual inattention is the orientation of the driver’s glances towards potential hazards via cuing. This method is based on the assumption that drivers are able to rapidly identify hazards once their attention is drawn to the area of interest regardless of preceding EOR duration. This work examined this assumption in a simulated automated driving context by projecting hazardous and nonhazardous road scenes to a participant while sitting in a stationary vehicle.

The electric vehicle on-board charger (OBC) is responsible for converting AC grid energy to DC energy to charge the battery pack. This paper describes the development of GM’s second generation OBC used in the 2016 Chevrolet Volt. The second generation OBC provides significant improvements in efficiency, size, and mass compared to the first generation. Reduced component count supports goals of improved reliability and lower cost. Complexity reduction of the hardware and diagnostic software was undertaken to eliminate potential failures.

The Computer-Aided Engineering of Automotive Batteries (CAEBAT) Phase 1 project is a U.S. Department of Energy-funded, multi-year project which is aimed at developing a complete CAE tool set for the automotive battery pack design. This paper reports the application of the full field approach of the CAEBAT which is developed by the General Motors-led industry team, for a 24-cell liquid-cooled prototype battery pack. It also summarizes the verification of the approach by comparing the simulation results with the measurement data. The simulation results using the Full Field Approach are found to have a very good agreement with the measurement data.

This paper describes the capabilities of a new two-motor plug-in hybrid-electric propulsion system developed for rear wheel drive. The PHEV system comprises a 2.0L turbocharged 4-cylinder direct-injected gasoline engine with the new hybrid transmission [1], a new traction power inverter module, a liquid-cooled lithium-ion battery pack, and on-board battery charger and 12V power converter module. The capability and features of the system components are described, and component performance and vehicle data are reported. The resulting propulsion system provides an excellent combination of electric-only driving, acceleration, and fuel economy.

Verification and validation (V&V) are essential stages in the design cycle of industrial controllers to remove the gap between the designed and implemented controller. In this study, a model-based adaptive methodology is proposed to enable easily verifiable controller design based on the formulation of a sliding mode controller (SMC). The proposed adaptive SMC improves the controller robustness against major implementation imprecisions including sampling and quantization. The application of the proposed technique is demonstrated on the engine cold start emission control problem in a mid-size passenger car. The cold start controller is first designed in a single-input single-output (SISO) structure with three separate sliding surfaces, and then is redesigned based on a multiinput multi-output (MIMO) SMC design technique using nonlinear balanced realization.

Test and verification procedures are a vital aspect of the development process for embedded control systems in the automotive domain. Formal requirements can be used in automated procedures to check whether simulation or experimental results adhere to design specifications and even to perform automatic test and formal verification of design models; however, developing formal requirements typically requires significant investment of time and effort for control software designers. We propose Signal Template Library (ST-Lib), a uniform modeling language to encapsulate a number of useful signal patterns in a formal requirement language with the goal of facilitating requirement formulation for automotive control applications. ST-Lib consists of basic modules known as signal templates. Informally, these specify a characteristic signal shape and provide numerical parameters to tune the shape.

The strain-induced diffusionless shear transformation of retained austenite to martensite during straining of transformation induced plasticity (TRIP) assisted steels increases strain hardening and delays necking and fracture leading to exceptional ductility and strength, which are attractive for automotive applications. A novel technique that provides the retained austenite volume fraction variation with strain with improved precision is presented. Digital images of the gauge section of tensile specimens were first recorded up to selected plastic strains with a stereo digital image correlation (DIC) system. The austenite volume fraction was measured by synchrotron X-ray diffraction from small squares cut from the gage section. Strain fields in the squares were then computed by localizing the strain measurement to the corresponding region of a given square during DIC post-processing of the images recorded during tensile testing.

Mahalanobis Distance (MD) is gaining momentum in many fields where classification, statistical pattern recognition, and forecasting are primary focus. It is a multivariate method and considers correlation relationships among parameters for computing generalized distance measure to separate groups or populations. MD is a useful statistic in multivariate analysis to test that an observed random sample is from a multivariate normal distribution. This capability alone enables engineers to determine if an observed sample is an outlier (defect) that falls outside the constructed (good) multivariate normal distribution. In Mahalanobis-Taguchi System (MTS), MD is suitably scaled and used as a measure of severity in abnormality assessment. It is obvious that computed MD depends on values of parameters observed on a random sample. All parameters may not equally impact MD. MD could be highly sensitive with respect to some parameters and less sensitive to some other parameters.