Search

For large-scale 3D point clouds from surveying, the RandLA-Net semantic segmentation network model is unable to learn global context features efficiently during training, leading to suboptimal feature acquisition of globally related features. Furthermore, inconsistent predictions are made in the vicinity of data boundaries due to the oversimplified usage of multi-layer perceptrons (MLPs) and non-linear activation functions following decoding, which has a detrimental impact on both model training and performance. This study provides two solutions to address these issues: a neighborhood label aggregation (NLA) classifier that aggregates neighborhood information for segmentation tasks, and a global context learning (GCL) module that learns global volume-relative information to improve the semantic segmentation network. Overall accuracy (OA) obtained via experimental validation on large-scale SensatUrban and S3DIS datasets is 91.80% and 88.8%, respectively. The proposed model outperforms RandLA-Net by 3.8% and 3.0% in overall mean Intersection over Union (mIOU), respectively, significantly improving model performance and generalization abilities. This work offers new perspectives on how to effectively segment point cloud data. [ABSTRACT FROM AUTHOR]

Understanding the impact of road network configuration, traffic stations and people flow factors on hotel layout can provide a scientific evidence for the spatial planning of hotels and other service industries. Based on multiple geographic big data, this study uses space syntax and Geographic Information System (GIS) to explore the impact of road network on the spatial layout of hotels with varying grades. Results show that low-end hotels are most affected by road network configurations; this is followed by mid-range hotels, and then by high-end hotels; high-end hotels are mostly located in the areas with high accessibility to the road network, while the mid-range and low-end hotels are mostly distributed near areas with strong traffic carrying capacity and attractiveness. Furthermore, traffic stations and people flow also have substantial impact on the spatial distribution of hotels, with the low-end hotels being most affected and the high-end hotels being least affected. Our findings provide several important implications for the spatial planning of hotels in the rapidly developing tourism industry. [ABSTRACT FROM AUTHOR]

In this paper, a waxberry‐like MnS/Ni3S4 composite catalyst was designed and synthesized. In this coating structure, MnS is located inside and Ni3S4 is wrapped on the surface to form a spherical structure. This structure makes the material show excellent stability in the electrocatalytic process. The diffusion staggered region structure formed at the two‐phase interface greatly enhances the synergistic interaction between MnS and Ni3S4. At the same time, the defects and vacancies formed by the diffusion mechanism at the interface of the two phases increase the active site and improve the interfacial electron transfer kinetics. Therefore, MnS/Ni3S4 composites showed good catalytic performance for ORR/OER. At 10 mA cm−2, the overpotential of it is only 320 mV, and the half‐wave potential can reach 0.81 V. The catalyst also exhibited extraordinary cycle stability and small voltage gap when used as cathode of Zn‐air batteries. When the current density is 3 mA cm−2, the cyclic discharge can be stable for 400 h and the voltage difference between the front and back does not increase more than. When the current density increases to 5 mA cm−2, the cyclic charge and discharge can be stable for more than 300 h. [ABSTRACT FROM AUTHOR]

Refined product distribution is an application of the Multi-compartment Vehicle Routing Problem (MCVRP), which simultaneously considers the vehicle routing, the assignment of heterogonous vehicles and loading policies of multi-compartment. First, we develop an optimization model with the objective of delivering on-time and minimizing transportation cost. Then we propose a Multi-objective Variable Neighborhood Descent Algorithm (MOVND), where the -constraint method transforms the original problem into a series of sub-problems of single objective with constraints. Finally, the efficiency of the proposed algorithm is verified by conducting a large number of small and large instances. Mainly including (i) compared with the classical NSGA-2 algorithm for multi-objective VRP, MOVND provides better performance in terms of convergence, spread and distribution; (ii) the multi-compartment vehicles are able to carry a variety of products simultaneously, which can improve the effective utilization of vehicle space and meet the needs of different customers in a single transportation. In addition, heterogonous vehicles that take full advantage of the characteristics of different vehicles are superior to homogeneous ones in terms of operating cost in the practice of the refined product distribution. [ABSTRACT FROM AUTHOR]

In this paper, error analysis of the unstructured mesh Galerkin finite element method for the two-dimensional time-space fractional Schrödinger equation with a time-independent potential defined on a finite domain is studied. The finite difference method is used to discretize the Caputo time fractional derivative, while the finite element method using unstructured mesh is used to deal with the Riesz fractional operators in space. Both the stability and convergence analysis of the numerical scheme are constructed. Numerical example is conducted to testify the validity of the proposed method. The conservation of the space fractional Schrödinger equation and the non-conservation of the time fractional Schrödinger equation in quantum mechanical system are achieved. This paper proposes an efficient numerical method as well as its theoretical analysis for the two-dimensional time-space fractional Schrödinger equation with time-independent potentials. [ABSTRACT FROM AUTHOR]