Search Results

The transition from combustion engines to electric propulsion is accelerating in every coordinate of the globe. The engineers had strived hard to augment the engine performance for more than eight decades, and a similar challenge had emerged again for electric vehicles. To analyze the performance of the engine, the vector engine operating point (EOP) is defined, which is common industry practice, and the performance vector electric vehicle motor operating point (EVMOP) is not explored in the existing literature. In an analogous sense, electric vehicles are embedded with three primary components, e.g., Battery, Inverter, Motor, and in this article, the EVMOP is defined using the parameters [motor torque, motor speed, motor current]. As a second aspect of this research, deep learning models are developed to predict the EVMOP by mapping the parameters representing the dynamic state of the system in real-time.

The automotive industry widely accepted the launch of electric vehicles in the global market, resulting in the emergence of many new areas, including battery health, inverter design, and motor dynamics. Maintaining the desired thermal stress is required to achieve augmented performance along with the optimal design of these components. The HVAC system controls the coolant and refrigerant fluid pressures to maintain the temperatures of [Battery, Inverter, Motor] in a definite range. However, identifying the prominent factors affecting the thermal stress of electric vehicle components and their effect on temperature variation was not investigated in real-time. Therefore, this article defines the vector electric vehicle thermal operating point (EVTHOP) as the first step with three elements [instantaneous battery temperature, instantaneous inverter temperature, instantaneous stator temperature].

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.

In the automotive industry, a Malfunction Indicator Light (MIL) is commonly employed to signify a failure or error in a vehicle system. To identify the root cause that has triggered a particular fault, a technician or engineer will typically run diagnostic tests and analyses. This type of analysis can take a significant amount of time and resources at the cost of customer satisfaction and perceived quality. Predicting an impending error allows for preventative measures or actions which might mitigate the effects of the error. Modern vehicles generate data in the form of sensor readings accessible through the vehicle’s Controller Area Network (CAN). Such data is generally too extensive to aid in analysis and decision making unless machine learning-based methods are used. This paper proposes a method utilizing a recurrent neural network (RNN) to predict an impending fault before it occurs through the use of CAN data.

Electrification is the well-accepted solution to address carbon emissions and modernize vehicle controls. Batteries play a critical in the journey of electrification and modernization with battery voltage prediction as the foundation for safe and efficient operation. Due to its strong dependency on prior information, battery voltage was estimated with recurrent neural network methods in the recent literatures exploring a variety of deep learning techniques to estimate battery behaviors. In these studies, standard recurrent neural networks, gated recurrent units, and long-short term memory are popular neural network architectures under review. However, in most cases, each neural network architecture is individually assessed and therefore the knowledge about comparative study among three neural network architecture is limited. In addition, many literatures only studied either the dynamic voltage response or the voltage relaxation.

This work presents a method to estimate the human body orientation using 2D images from a person view; the challenge comes from the variety of human body poses and appearances. The method utilizes OpenPose neural network as a human pose detector module and depth sensing module. The modules work together to extract the body orientation from 2D stereo images. OpenPose is proven to be efficient in detecting human body joints, defined by COCO dataset, OpenPose can detect the visible body joints without being affected by backgrounds or other challenging factors. Adding the depth data for each point can produce rich information to the process of 3D construction for the detected humans. This 3D point’s setup can tell more about the body orientation and walking direction for example. The depth module used in this work is the ZED camera stereo system which uses CUDA for high performance depth computation.

The Press-in-Place (PIP) gasket is a static face seal with self-retaining feature, which is used for the mating surfaces of engine components to maintain the reliability of the closed system under various operating conditions. Its design allows it to provide enough contact pressure to seal the internal fluid as well as prevent mechanical failures. Insufficient sealing pressure will lead to fluid leakage, consequently resulting in engine failures. A test fixture was designed to simulate the clamp load and internal pressure condition on a gasket bolted joint. A sensor pad in combination with TEKSCAN equipment was used to capture the overall and local pressure distribution of the PIP gasket under various engine loading conditions. Then, the test results were compared with simulated results from computer models. Through the comparisons, it was found that gasket sealing pressure of test data and CAE data shows good correlations in all internal pressure cases when the bolt load was 500 N.

As the industry moves further into the automotive age, the failure of the cup during the transportation of the parts during the assembly process is costly. Among them, the effective contact area of the suction cup could influence the significant availability of the pressure, which is necessary to investigate the truth. The essential objective for this research is trying to improve the effectiveness of the suction cups during gripers work in company’s industry. In this research, the real work condition is simulated by the experimental setup to find the influence of the effective contact area. In this paper, the proper methodology to measure the effective area by testing different size cups under different conditions is described. The results are verified by the digital image correlation (DIC) technique.

An experimental investigation of non-intrusive combustion sensing was performed using a tri-axial accelerometer mounted to the engine block of a small-bore high-speed 4-cylinder compression-ignition direct-injection (CIDI) engine. This study investigates potential techniques to extract combustion features from accelerometer signals to be used for cycle-to-cycle engine control. Selection of accelerometer location and vibration axis were performed by analyzing vibration signals for three different locations along the block for all three of the accelerometer axes. A magnitude squared coherence (MSC) statistical analysis was used to select the best location and axis. Based on previous work from the literature, the vibration signal filtering was optimized, and the filtered vibration signals were analyzed. It was found that the vibration signals correlate well with the second derivative of pressure during the initial stages of combustion.

This work presents a method for estimating human body orientation using a combination of convolutional neural network (CNN) and stereo camera in real time. The approach uses the CNN model to predict certain human body keypoints then transforms these points into a 3D space using the stereo vision system to estimate the body orientations. The CNN module is trained to estimate the shoulders, the neck and the nose positions, detecting of three points is required to confirm human detection and provided enough data to translate the points into 3D space.

The article solves the problem of reducing electric power losses of the traction induction machine rotor to prevent its overheating in nominal and high-load modes. Electric losses of the rotor power are optimized by the stabilization of the main magnetic flow of the electric machine at a nominal level with the amplitude-frequency control in a wide range of speeds and increased loads. The quasi-independent excitation of the induction machine allows us to increase the rigidity of mechanical characteristics, decrease the rotor slip at nominal loads and overloads and significantly decrease electrical losses in the rotor as compared to other control methods. The article considers the technology of converting the power of individual phases into a single energy flow using a three-phase electric machine equivalent circuit and obtaining an energy model in the form of equations of instantaneous active and reactive power balance.

An existing challenge in current state-of-the-art autonomous vehicles is the process of safely transferring control from autonomous driving mode to manual mode because the driver may be distracted with secondary tasks. Such distractions may impair a driver’s situational awareness of the driving environment which will lead to fatal outcomes during a handover. Current state-of-the-art vehicles notify a user of an imminent handover via auditory, visual, and physical alerts but are unable to improve a driver’s situational awareness before a handover is executed. The overall goal of our research team is to address the challenge of providing a driver with relevant information to regain situational awareness of the driving task. In this paper, we introduce a novel approach to estimating a driver’s visual focus of attention using a 2D RGB camera as input to a Multi-Input Convolutional Neural Network with shared weights. The system was validated in a realistic driving scenario.

This paper presents an experimental investigation of the impact of EGR dilution on the tradeoff between flame and end-gas autoignition heat release in a Spark-Assisted Compression Ignition (SACI) combustion engine. The mixture was maintained stoichiometric and fuel-to-charge equivalence ratio (ϕ′) was controlled by varying the EGR dilution level at constant engine speed. Under all conditions investigated, end-gas autoignition timing was maintained constant by modulating the mixture temperature and spark timing. Experiments at constant intake pressure and constant spark timing showed that as ϕ′ is increased, lower mixture temperatures are required to match end-gas autoignition timing. Higher ϕ′ mixtures exhibited faster initial flame burn rates, which were attributed to the higher laminar flame speeds immediately after spark timing and their effect on the overall turbulent burning velocity.

This work analyzes a cantilevered piezoelectric beam device for harvesting energy from the simultaneous rotation and translational vibration of vehicle wheels. The device attaches to the wheel rim so that it displaces tangentially during operation. A lumped-parameter analytical model for the coupled electromechanical system is derived. The device has one natural frequency that is speed-dependent because of centripetal acceleration affecting the total stiffness of the device. Even though the device has one natural frequency, it experiences three resonances as the rotation speed varies. One resonance occurs when the rotation speed coincides with the speed-dependent natural frequency of the device. The other two resonances are associated with excitations from the vibration of the vehicle wheel. The device’s parameters are chosen so that these three resonances occur when the wheel travels near 30 mph, 55 mph, and 70 mph.

Machine learning methods, such as decision trees and deep neural networks, are becoming increasingly important and useful for data analysis in various scientific fields including dynamics and control, signal processing, pattern recognition, fluid mechanics, and chemical synthesis, etc. For future engine design and performance optimization, there is an urgent need for a robust predictive model which could capture the major combustion properties such as autoignition and flame propagation of multicomponent fuels under a wide range of engine operating conditions, without massive experimental measurement or computational efforts. It will be shown that these long-held limitations and challenges related to complex fuel combustion and engine research could be readily solved by implementing machine learning methods.

Flame speed data reported in most literature are acquired in conventional apparatus such as the spherical combustion bomb and counterflow burner, and are limited to atmospheric pressure and ambient or slightly elevated unburnt temperatures. As such, these data bear little relevance to internal combustion engines and gas turbines, which operate under typical pressures of 10-50 bar and unburnt temperature up to 900K or higher. These elevated temperatures and pressures not only modify dominant flame chemistry, but more importantly, they inevitably facilitate pre-ignition reactions and hence can change the upstream thermodynamic and chemical conditions of a regular hot flame leading to modified flame properties. This study focuses on how auto-ignition chemistry affects flame propagation, especially in the negative-temperature coefficient (NTC) regime, where dimethyl ether (DME), n-heptane and iso-octane are chosen for study as typical fuels exhibiting low temperature chemistry (LTC).

There is a pressing need to develop accurate and robust approaches for predicting vehicle speed to enhance fuel economy/energy efficiency, drivability and safety of automotive vehicles. This paper details outcomes of research into various methods for the prediction of vehicle velocity. The focus is on short-term predictions over 1 to 10 second prediction horizon. Such short-term predictions can be integrated into a hybrid electric vehicle energy management strategy and have the potential to improve HEV energy efficiency. Several deterministic and stochastic models are considered in this paper for prediction of future vehicle velocity. Deterministic models include an Auto-Regressive Moving Average (ARMA) model, a Nonlinear Auto-Regressive with eXternal input (NARX) shallow neural network and a Long Short-Term Memory (LSTM) deep neural network. Stochastic models include a Markov Chain (MC) model and a Conditional Linear Gaussian (CLG) model.

As European and Chinese tailpipe emission regulations for gasoline light-duty vehicles impose particulate number limits, automotive manufacturers have begun equipping some vehicles with a gasoline particulate filter (GPF). Increased understanding of how soot morphology, reactivity, and GPF loading affect GPF filtration and regeneration characteristics is necessary for advancing GPF performance. This study investigates the impacts of morphology, reactivity, and filter soot loading on GPF filtration and regeneration. Soot morphology and reactivity are varied through changes in fuel injection parameters, known to affect soot formation conditions. Changes in morphology and reactivity are confirmed through analysis using a transmission electron microscope (TEM) and a thermogravimetric analyzer (TGA) respectively.

Combined direct and port fuel injection (i.e., dual injection) in spark ignition engines is of increasing interest due to the advantages for fuel flexibility and the individual merits of each system for improving engine performance and reducing engine-out emissions. Greater understanding of the impact of dual injection will enable deriving the maximum benefit from the two injection systems. This study investigates the effects of dual injection on combustion, especially knock propensity and tolerance to exhaust gas recirculation (EGR) dilution at different levels of EGR. A baseline for comparison with dual injection results was made using direct injection fueling only. A splash blended E20 fuel was used for the direct injection only tests. For the dual injection tests, gasoline, representing 80% by volume of the total fuel, was injected using the direct injector, and ethanol, representing 20% by volume of the total fuel, was injected using the port fuel injector.

Non-reacting spray behaviors of the Engine Combustion Network “Spray G” gasoline fuel injector were investigated at flash and non-flash boiling conditions in an optically accessible single cylinder engine and a constant volume spray chamber. High-speed Mie-scattering imaging was used to determine transient liquid-phase spray penetration distances and observe general spray behaviors. The standardized “G2” and “G3” test conditions recommended by the Engine Combustion Network were matched in this work and the fuel was pure iso-octane. Results from the constant volume chamber represented the zero (stationary piston) engine speed condition and single cylinder engine speeds ranged from 300 to 2,000 RPM. As expected, the present results indicated the general spray behaviors differed significantly between the spray chamber and engine. The differences must be thoughtfully considered when applying spray chamber results to guide spray model development for engine applications.