Compression ratio

Abstract: The development of four-stroke, spark-ignition engines that are designed to inject gasoline directly into the combustion chamber is an important worldwide initiative of the automotive industry. The thermodynamic potential of such engines for significantly enhanced fuel economy, transient response and cold-start hydrocarbon emission levels has led to a large number of research and development projects that have the goal of understanding, developing and optimizing gasoline direct-injection (GDI) combustion systems. The processes of fuel injection, spray atomization and vaporization, charge cooling, mixture preparation and the control of in-cylinder air motion are all being actively researched, and this work is reviewed in detail and analyzed. The new technologies such as high-pressure, common-rail, gasoline injection systems and swirl-atomizing gasoline fuel injectors are discussed in detail, as these technologies, along with computer control capabilities, have enabled the current new examination of an old objective; the direct-injection, stratified-charge (DISC), gasoline engine. The prior work on DISC engines that is relevant to current GDI engine development is also reviewed and discussed. The fuel economy and emission data for actual engine configurations are of significant importance to engine researchers and developers. These data have been obtained and assembled for all of the available GDI literature, and are reviewed and discussed in detail. The types of GDI engines are arranged in four classifications of decreasing complexity, and the advantages and disadvantages of each class are noted and explained. Emphasis is placed upon consensus trends and conclusions that are evident when taken as a whole. Thus the GDI researcher is informed regarding the degree to which engine volumetric efficiency and compression ratio can be increased under optimized conditions, and as to the extent to which unburned hydrocarbon (UBHC), NOx and particulate emissions can be minimized for specific combustion strategies. The critical area of GDI fuel injector deposits and the associated effect on spray geometry and engine performance degradation are reviewed, and important system guidelines for minimizing deposition rates and deposit effects are presented. The capabilities and limitations of emission control techniques and aftertreatment hardware are reviewed in depth, and areas of consensus on attaining European, Japanese and North American emission standards are compiled and discussed. All known research, prototype and production GDI engines worldwide are reviewed as to performance, emissions and fuel economy advantages, and for areas requiring further development. The engine schematics, control diagrams and specifications are compiled, and the emission control strategies are illustrated and discussed. The influence of lean-NOx catalysts on the development of late-injection, stratified-charge GDI engines is reviewed, and the relative merits of lean-burn, homogeneous, direct-injection engines as an option requiring less control complexity are analyzed. All current information in the literature is used as the basis for discussing the future development of automotive GDI engines.

Abstract: Fractals are geometric or data structures which do not simplify under magnification. Fractal Image Compression is a technique which associates a fractal to an image. On the one hand, the fractal can be described in terms of a few succinct rules, while on the other, the fractal contains much or all of the image information. Since the rules are described with less bits of data than the image, compression results. Data compression with fractals is an approach to reach high compression ratios for large data streams related to images. The high compression ratios are attained at a cost of large amounts of computation. Both lossless and lossy modes are supported by the technique. The technique is stable in that small errors in codes lead to small errors in image data. Applications to the NASA mission are discussed.

Abstract: Large-scale distributed training requires significant communication bandwidth for gradient exchange that limits the scalability of multi-node training, and requires expensive high-bandwidth network infrastructure. The situation gets even worse with distributed training on mobile devices (federated learning), which suffers from higher latency, lower throughput, and intermittent poor connections. In this paper, we find 99.9% of the gradient exchange in distributed SGD is redundant, and propose Deep Gradient Compression (DGC) to greatly reduce the communication bandwidth. To preserve accuracy during compression, DGC employs four methods: momentum correction, local gradient clipping, momentum factor masking, and warm-up training. We have applied Deep Gradient Compression to image classification, speech recognition, and language modeling with multiple datasets including Cifar10, ImageNet, Penn Treebank, and Librispeech Corpus. On these scenarios, Deep Gradient Compression achieves a gradient compression ratio from 270x to 600x without losing accuracy, cutting the gradient size of ResNet-50 from 97MB to 0.35MB, and for DeepSpeech from 488MB to 0.74MB. Deep gradient compression enables large-scale distributed training on inexpensive commodity 1Gbps Ethernet and facilitates distributed training on mobile.

Abstract: A premixed charge compression ignition engine, and a control system, is provided which effectively initiates combustion by compression ignition and maintains stable combustion while achieving extremely low nitrous oxide emissions, good overall efficiency and acceptable combustion noise and cylinder pressures. The present engine and control system effectively controls the combustion history, that is, the time at which combustion occurs, the rate of combustion, the duration of combustion and/or the completeness of combustion, by controlling the operation of certain control variables providing temperature control, pressure control, control of the mixture's autoignition properties and equivalence ratio control. The combustion control system provides active feedback control of the combustion event and includes a sensor, e.g. pressure sensor, for detecting an engine operating condition indicative of the combustion history, e.g. the start of combustion, and generating an associated engine operating condition signal. A processor receives the signal and generates control signals based on the engine operating condition signal for controlling various engine components to control the temperature, pressure, equivalence ratio and/or autoignition properties so as to variably control the combustion history of future combustion events to achieve stable, low emission combustion in each cylinder and combustion balancing between the cylinders.

Abstract: Low temperature combustion (LTC) engines are an emerging engine technology that offers an alternative to spark-ignited and diesel engines. One type of LTC engine, the homogeneous charge compression ignition (HCCI) engine, uses a well-mixed fuel–air charge like spark-ignited engines and relies on compression ignition like diesel engines. Similar to diesel engines, the use of high compression ratios and removal of the throttling valve in HCCI allow for high efficiency operation, thereby allowing lower CO 2 emissions per unit of work delivered by the engine. The use of a highly diluted well-mixed fuel–air charge allows for low emissions of nitrogen oxides, soot and particulate matters, and the use of oxidation catalysts can allow low emissions of unburned hydrocarbons and carbon monoxide. As a result, HCCI offers the ability to achieve high efficiencies comparable with diesel while also allowing clean emissions while using relatively inexpensive aftertreatment technologies. HCCI is not, however, without its challenges. Traditionally, two important problems prohibiting market penetration of HCCI are 1) inability to achieve high load, and 2) difficulty in controlling combustion timing. Recent research has significantly mitigated these challenges, and thus HCCI has a promising future for automotive and power generation applications. This article begins by providing a comprehensive review of the physical phenomena governing HCCI operation, with particular emphasis on high load conditions. Emissions characteristics are then discussed, with suggestions on how to inexpensively enable low emissions of all regulated emissions. The operating limits that govern the high load conditions are discussed in detail, and finally a review of recent research which expands the high load limits of HCCI is discussed. Although this article focuses on the fundamental phenomena governing HCCI operation, it is also useful for understanding the fundamental phenomena in reactivity controlled compression ignition (RCCI), partial fuel stratification (PFS), partially premixed compression ignition, spark-assisted HCCI, and all forms of low temperature combustion (LTC).