Search Results

Visual Place Recognition (VPR) excels at providing a good location prior for autonomous vehicles to initialize the map-based visual SLAM system, especially when the environment changes after a long term. Condition change and viewpoint change, which influences features extracted from images, are two of the major challenges in recognizing a visited place. Existing VPR methods focus on developing the robustness of global feature to address them but ignore the benefits that local feature can auxiliarily offer. Therefore, we introduce a novel hierarchical place recognition method with both global and local features deriving from homologous VLAD to improve the VPR performance. Our model is weak supervised by GPS label and we design a fine-tuning strategy with a coupled triplet loss to make the model more suitable for extracting local features.

The front camera module is a fundamental component of a modern vehicle’s active safety architecture. The module supports many active safety features. Perception of the road environment, requests for driver notification or alert, and requests for vehicle actuation are among the camera software’s key functions. This paper presents a novel method of testing these functions virtually. First, the front camera module software is compiled and packaged in a Docker container capable of running on a standard Linux computer as a software in the loop (SiL). This container is then integrated with the active safety simulation tool that represents the vehicle plant model and allows modeling of test scenarios. Then the following simulation components form a closed loop: First, the active safety simulation tool generates a video data stream (VDS). Using an internet protocol, the tool sends the VDS to the camera SiL and other vehicle channels.

A test cell was developed for evaluating a 6-speed automatic transmission. The target vehicle had an internal combustion 5.4L gasoline V8 engine. An electric dynamometer was used to closely simulate the engine characteristics. This included generating mean torque from the ECU engine map, with a transient capability of 10,000 rpm/second. Engine inertia was simulated with a transient capability of 20,000 rpm/second, and torque pulsation was simulated individually for each piston, with a transient capability of 50,000 rpm/second. Quantitative results are presented for the correlation between the engine driven and the dynamometer driven transmission performance over more than 60 test cycles. Concerns about using the virtual engine in validation testing are discussed, and related to the high frequency transient performance required from the electric dynamometer. Qualitative differences between the fueled engine and electric driven testing are presented.

The feasibility of using experimentally determined flow fields at intake valve closing, IVC, as initial conditions for computing the in-cylinder flow dynamics during the compression stroke is demonstrated by means of a computer simulation of the overall approach. A commercial CFD code, STAR-CD, was used for this purpose. The study involved two steps. First, in order to establish a basis for comparison, the in-cylinder flow field throughout the intake and compression strokes, from intake valve opening, IVO, to top dead center, TDC, was computed for a simple engine geometry. Second, experimental initial conditions were simulated by randomly selecting and perturbing a set of velocity vectors from the computed flow field at IVC.

With the development of economic globalization, the speed of development of container terminals is also very rapid. Under the pressure brought by the surge in throughput, the unmanned and intelligent terminals will become the future development direction of terminals. As the cornerstone of the unmanned terminal, the positioning technology provides the most basic position information for system scheduling, path planning, real-time correction, and loading and unloading. Therefore, this paper is aimed to design a low-cost, high-precision, and easy-to-maintain unmanned dock positioning system in order to better solve the problem of unmanned dock positioning. The main research content of this paper is to design a positioning algorithm for unmanned terminal Automated Guided Vehicle (AGV) based on single-line lidar, including point cloud data acquisition, background filtering, point cloud clustering, vehicle position extraction, and result optimization.

Rotating flow inside an internal combustion engine cylinder is deliberately engineered for improved fuel-air mixing and combustion. The details of the rotating flow structure vary temporally over an engine cycle as well as cyclically at the same engine phase. Algorithms in the literature to identify these structural details of the rotating flow invariably focus on locating its center and, on occasion, measuring its rotational strength and spatial extent. In this paper, these flow structure parameters are evaluated by means of complex moments, which have been adapted from image (scalar field) recognition applications to two-dimensional flow pattern (vector field) analysis. Several additional detailed characteristics of the rotating flow pattern - the type and extent of its deviation from the ideal circular pattern, its rotational and reflectional symmetry (if exists), and thus its orientation - are also shown to be related to the first few low-order complex moments of the flow pattern.

Gas carburized and quenched low alloy steels typically produce surface microstructures which contain martensite, retained austenite and often NMTP's (non-martensitic transformation products). The NMTP's are caused by a reduction of surface hardenability in the carburizing process from loss of alloying elements to oxidation. Gas carburized low alloy steels such as SAE 8620 with NMTP's on the surface have been shown to have inferior bending fatigue properties when compared to more highly alloyed steels which do not form NMTP's, such as SAE 4615M. One method of minimizing the formation of oxides and eliminating NMTP formation during carburizing and quenching is to use plasma carburizing instead of conventional gas carburizing. In this study the microstructures and bending fatigue performance of plasma carburized SAE 8620 and SAE 4615M is compared to the same alloys conventionally gas carburized and quenched.

Application of steady-flow correlations to characterize flow losses in complex piping systems is well established for non-transient fluid transport engineering. As a result, the literature contains numerous correlations relating flow (or pressure) losses to the piping system geometry. The present study applies these correlations to an intake manifold of a four cylinder engine to identify regions in the manifold that contribute most significantly to the system flow loss; results showed that the primary runner entrances accounted for over half of the total system loss. With this finding, four manifolds were designed and tested on a steady-flow bench and on an engine. Reduced flow losses resulted in improved peak engine performance at the expense of low speed volumetric efficiency. Primary runner pressures at peak performance conditions were analyzed in both the time and frequency domain.

A process for creating a Customer Correlated, Accelerated, Life Test is presented. This process, which results in a model for predicting reliability, has been applied to a cold weather piston scuff problem. In this paper, the authors will discuss development of frequency distributions for customer environmental and operational use, establishment of customer based failure criteria, development of an accelerated test based on degradation, selection of testing strategies, data analyses, and measurement techniques.

Flash boiling spray has been proven to be a useful method in providing finer fuel droplet and stronger evaporation in favor of creating a homogeneous fuel-air mixture. Combustion characteristics of flash boiling spray are thus valuable to be investigated systematically for aiding the development of efficient internal combustion system. An experimental study of flash boiling spray combustion in a SIDI optical engine under early injection has been conducted. The fuel, Iso-octane, was used across all tests. Three fuel spray conditions experimented in the study: normal liquid, transitional flash boiling and flare flash boiling sprays, within each case that Pa/Ps ratio was set in (>1), (0.3~1), and (<0.3) respectively. A small quartz insert on the piston enables optical access for observing combustion process; non-intrusive measurements on flame radicals has been carried out using a high-speed color camera.

Spark knock pressure oscillations can be detected by a cylinder pressure transducer or by an accelerometer mounted on the engine block. Accelerometer-based detection is lower cost but is affected by extraneous mechanical vibrations and the frequency response of the engine block and accelerometer. The knock oscillation frequency changes during the expansion stroke because the chamber geometry is changing due to the piston motion and the burned gases are cooling. Spectrogram analysis shows the time-dependent frequency content of the pressure and acceleration signals, revealing characteristic signatures of knock and mechanical vibrations. Illustrative spectrograms are presented which yield physical insight into accelerometer-based knock detection.

Torque converter clutch friction interface and automatic transmission fluid (ATF) temperatures, pressure difference across the clutch piston, flow through the friction material grooves, and engine crankshaft dynamic torque were measured for typical operating conditions on a running transmission. The friction coefficient, clutch unit pressure, fraction of heat rejected to ATF flowing through the grooves, and time dependent thermal response were determined. Simplified heat transfer calculations were compared with thermal data. Clutch interface temperatures were assessed as they relate to the process of friction material and ATF degradation. The inductively powered radiotelemetry system was found to be a robust and powerful tool for investigating continuously slipping clutch system performance.

Lift-off length is defined as the distance from injector hole to the location where flame stabilized on a high injection pressure direct injection (DI) diesel spray. In this paper we used the high-speed (40 kHz) Schlieren and time-averaged OH chemiluminescence imaging technique to simultaneously measure the flame lift-off locations on a DI diesel spray in an optically accessible and constant-volume combustion vessel. The time-resolved development of the diesel spray acquired from the high-speed Schlieren imaging system enabled us to observe the instantaneous spray structure details of the spray flames. The OH chemiluminescence image obtained from a gated, intensified CCD video camera with different delay and width settings was used to determine the quiescent lift-off length. Experiments were conducted under various ambient temperatures, ambient gas densities, injection pressures and oxygen concentrations.

This work reports a CFD study on a 2-stroke (2-S) opposed piston high speed direct injection (HSDI) Diesel engine. The engine main features (bore, stroke, port timings, et cetera) are defined in a previous stage of the project, while the current analysis is focused on the assembly made up of scavenge ports, manifold and cylinder. The first step of the study consists in the construction of a parametric mesh on a simplified geometry. Two geometric parameters and three different operating conditions are considered. A CFD-3D simulation by using a customized version of the KIVA-4 code is performed on a set of 243 different cases, sweeping all the most interesting combinations of geometric parameters and operating conditions. The post-processing of this huge amount of data allow us to define the most effective geometric configuration, named baseline.

Prior to making rotational vibration measurements with a laser vibrometer, it is good practice to establish that the instrument is operating properly. This can be accomplished by comparative measurement of a rotational vibration source with known amplitude and frequency. This paper describes the design and development of a rotational vibration apparatus with known amplitude and frequency to be used as a reference for comparison to concurrent and co-located measurements made by a rotational laser vibrometer (RLV). The comparative measurements acquired with the apparatus are helpful to verify proper laser vibrometer operation in between regular calibration intervals, and/or whenever the functionality of the vibrometer is suspect. In the subject apparatus, a Cardan shaft with variable input speed and angle is used to provide output torsional vibration with variable frequency and amplitude.

Rheological measurements were performed on a series of lubricants for flexible fuel vehicles, and blends of water or methanol in these oils. A series of measurements, including kinematic viscosity, viscosity at low and high shear rates, low shear viscosity under borderline pumping conditions, and density were performed on all oils and blends. The effects of mixing conditions, such as mixing speed and temperature on these properties were also studied. Viscosity increases when water emulsifies in oils. Methanol exhibits limited solubility in all oils, but more so in synthetic base oils. Viscosity tests at 248 K (-25°C) do not indicate the onset of critical pumping conditions, even at high concentrations of water or methanol. Tests at high shear rates at 323 K (50°C) suggest that water-oil emulsions are quite stable, while methanol-oil blends lose their methanol content either due to evaporation or shear-induced separation.

Aiming at the high altitude operation problems for piston-type aero-engines and to improve the practical ceiling and high altitude dynamic performance, this thesis analyzes a controllable three-stage composite supercharging system, using a two-stage turbocharger coupled supercharger method. The GT-Power simulation model of a four-cylinder boxer engine was established, and the control strategy of variable flight height was obtained. The simulation research of engine performance from 0 to 20,000 meters above sea level has been carried out, which shows that the engine power is at the same level as the plain condition, and it could still maintain 85.28 percent of power even at the height of 20,000 meters, which meets the flight requirements of the aircraft.

There is a clearly demonstrated need to develop “real-time” methods for the measurement of diesel vehicle particulate emissions. Optical techniques provide One alternative for such methodology because of the rapid data acquisition times involved and the relatively simple sampling methods that can be used. This report describes two different approaches to this problem. The first, the spectrophone, measures light absorption by the diluted exhaust plume using photoacoustic spectroscopy, and the second, the long pathlength smokemeter, determines total light extinction across the diluted exhaust plume. For the measurement of total mass emissions, both techniques show estimated errors of ±15% for specific vehicles, while for a multi-vehicle diesel fleet the estimated errors are ±20% and ±30% for the long pathlength smokemeter and the spectrophone, respectively.

A Machine Learning (ML) approach is presented to correlate in-cylinder images of early flame kernel development within a spark-ignited (SI) gasoline engine to early-, mid-, and late-stage flame propagation. The objective of this study was to train machine learning models to analyze the relevance of flame surface features on subsequent burn rates. Ultimately, an approach of this nature can be generalized to flame images from a variety of sources. The prediction of combustion phasing was formulated as a regression problem to train predictive models to supplement observations of early flame kernel growth. High-speed images were captured from an optically accessible SI engine for 357 cycles under pre-mixed operation. A subset of these images was used to train three models: a linear regression model, a deep Convolutional Neural Network (CNN) based on the InceptionV3 architecture and a CNN built with assisted learning on the VGG19 architecture.

Rapid Prototyping, Fabrication and Tooling is a process that blends a series of technologies (machines, tools, and methods) capable of generating physical objects directly from a CAD database. The process dramatically reduces the time spent during product development by allowing for fast visualization, verification, iteration, optimization, and fabrication of parts and tools. Many new techniques of tooling have been and are being developed by using rapid fabricated parts. These are having a dramatic impact on both timing and costs throughout the automotive industry. One area that these methods can be utilized to their full potential is motorsports. Of particular interest is the growing use of bridge tooling to provide first article through production intent parts that promote cost effective changes.