Physical properties of biodiesel play an important role in the injection, atomization and combustion performance. An experiment was carried out to investigate the spray properties of biodiesel. The experimental setup was based on an electronic unit-pump (EUP) bench, a constant volume chamber and a high-speed digital camera. The photographs of spray were dealt with by using an image processing procedure. Then the spray tip penetrations and cone angles were obtained and analyzed. The experimental results indicate that the spray tip penetrations and cone angles of biodiesel increase with increasing injection duration. In addition, with decreasing ambient pressure the spray tip penetrations increase while the cone angles decrease. Furthermore, ambient pressure has a stronger effect on the spray properties of biodiesel than injection pressure. On the macroscopically view, the shape of biodiesel spray is similar to that of diesel. The final tip penetrations and cone angles of biodiesel are greater than those of diesel due to its higher viscosity, density and bulk modulus. Copyright © 2008 John Wiley & Sons, Ltd.

Accurate real-time information on road slopes and the capacity to forecast future moment gradient values are critical for the vehicle control, stability, and driving comfort. Thus, this study proposes a stacking model method for road slope estimation of electric vehicles. Gated Circulation unit (GRU), Convolutional Neural Network (CNN), and CNN-GRU are used as the base classifiers, and Multilayer Perceptron (MLP) is used as a meta-classifier. The vehicle dynamics equations are examined to select the appropriate parameters to feed into the base classifier for training. The meta-classifier is trained using the estimated results from the basic classifier. The current slope values are estimated by slicing the training set by data sampling time and windowing the training data set to predict the future slope values in 2s, 3s, and 4s. Road experiments are conducted, and error indicators are selected for evaluation. The stacking model is compared with each base classifier, Adaptive Kalman filter, Recursive Least Squares with Forgetting Factor and Back Propagation Neural Network for estimating the current moment slope, and it is verified that the stacking model can better estimate the current slope value and outperform the conventional algorithm. Comparing the stacking model with the predicted results of each base classifier for future time slope prediction shows that the stacking model is more accurate at predicting the slope values in the short future time.

The geometric alignment of highways directly affects the fuel consumption of motor vehicles. To analyze the impact of the road gradient on fuel consumption, actual road tests were conducted in plateau and mountainous areas. Geographic Information Systems (GISs) were used to calculate road gradients, and Vehicle Specific Power (VSP) distributions were obtained through testing, serving as inputs for the motor vehicle emission simulator (MOVES) model. Finally, the simulation results were verified against the experimental results. The findings indicate a strong positive correlation between road gradients ranging from −5% to +5%, the VSP, and fuel consumption. At a constant gradient, the fuel consumption rate increases with the vehicle speed; the fuel consumption factor is lowest at 60 km/h and highest at 40 km/h. Under both constant and actual driving speeds, when the absolute values of uphill and downhill gradients are the same, the average fuel consumption for both uphill and downhill driving shows that, at gradients of 1% to 3%, the fuel savings from downhill driving can offset the additional fuel consumption on uphill driving. At gradients of 4% to 6%, the increase in fuel consumption on uphill driving surpasses the savings from downhill driving. During uphill climbs, lower speeds within the mid-to-low range result in lower fuel consumption and greater reserve power. The MOVES model demonstrates good adaptability in plateau and mountainous areas.

SAR ship detection faces critical deployment challenges: existing models exhibit excessive computational requirements and inadequate small-target performance for practical maritime surveillance. We propose LightSAR-Net, a lightweight multi-scale framework addressing these limitations through systematic architectural optimization.The network integrates three innovations: DEF-Stem employs dual-branch processing with Sobel edge enhancement to suppress sea clutter while preserving ship contours; VoV-GSCSP utilizes group-hybridized convolution with dynamic weighting for parameter-efficient feature fusion; DAT-DH decouples classification-regression tasks through spatial probability mapping and deformable convolution for enhanced dense target discrimination.SSDD evaluation achieves 97.1% mAP_0.5 (+1.0% over baseline) while reducing parameters by 29.9% and computation by 19.1%, demonstrating effective optimization for satellite edge computing platforms.