3D object detection with LiDAR and camera fusion has always been a challenge for autonomous driving. This work proposes a deep neural network (namely FuDNN) for LiDAR–camera fusion 3D object ...detection. Firstly, a 2D backbone is designed to extract features from camera images. Secondly, an attention-based fusion sub-network is designed to fuse the features extracted by the 2D backbone and the features extracted from 3D LiDAR point clouds by PointNet++. Besides, the FuDNN, which uses the RPN and the refinement work of PointRCNN to obtain 3D box predictions, was tested on the public KITTI dataset. Experiments on the KITTI validation set show that the proposed FuDNN achieves AP values of 92.48, 82.90, and 80.51 at easy, moderate, and hard difficulty levels for car detection. The proposed FuDNN improves the performance of LiDAR–camera fusion 3D object detection in the car category of the public KITTI dataset.
Two Approaches to Genre Analysis Xiao, Zhonghua; McEnery, Anthony
Journal of English linguistics,
03/2005, Letnik:
33, Številka:
1
Journal Article
Recenzirano
This article compares two approaches to genre analysis: Biber’s multidimensional analysis (MDA) and Tribble’s use of the keyword function of WordSmith. The comparison is undertaken via a case study ...of conversation, speech, and academic prose in modern American English. The terms conversation and speech as used in this article correspond to the demographically sampled and context-governed spoken data in the British National Corpus. Conversation represents the type of communication we experience every day whereas speech is produced in situations in which there are few producers and many receivers (e.g., classroom lectures, sermons, and political speeches). Academic prose is a typical formal-written genre that differs markedly from the two spoken genres. The results of the MDA and keyword approaches both on similar genres (conversation vs. speech) and different genres (the two spoken genres vs. academic prose) show that a keyword analysis can capture important genre features revealed by MDA.
Accurate energy consumption prediction models can bring tremendous benefits to building energy efficiency, where the use of data-driven models allows models to be trained based on historical data and ...to obtain better prediction performance. However, a severe problem has been overlooked where the dynamic spatiotemporal dependence among buildings is often not considered. To settle this issue, this study proposes an asymmetric encoder-decoder learning framework where the spatial relationships and time-series features between multiple buildings are extracted by a convolutional neural network and a gated recurrent neural network to form new input data in the encoder. The decoder then makes predictions based on the input data with an attention mechanism. The model is trained and tested with energy consumption data from 10 different buildings at a university in China. The results show that our proposed model maintains high accuracy using small datasets while predicting different types of buildings, achieving an average R2 of 0.964. Additionally, the accuracy loss in multi-step forecasting is considerably ameliorated by maintaining an average R2 of 0.915 predicting 3 steps ahead of time. Benefiting from the asymmetric encoder-decoder structure, the proposed algorithm resolves the problem of accuracy loss using deep learning in a small dataset and multi-step time series forecasting. Further assisted by the anomaly detection function, the algorithm serves as a reliable guide for the user to effectively manage and control energy consumption in buildings.
•An asymmetric encoder-decoder learning framework is proposed.•Clustering for anomaly building energy detection is included.•A comparison is made between the proposed model and other deep learning models.•High accuracy and generality are achieved in different buildings and small datasets.•The accuracy loss in multi-step forecasting is ameliorated.
•An intelligent fire emergency response analysis model incorporating HTCPN is proposed.•A SHAP sensitivity analysis is conducted to identify the relevant importance of key factors.•Busy rate of place ...and the utilization rate of transition are used to identify the key influence parts.•Uncertainties underlying complex coordination among various departments are considered.•The proposed model can significantly improve the efficiency of emergency response processes.
This paper proposes an intelligent fire emergency rescue analysis model in which the Hierarchical Timed Colored Petri nets (HTCPN) are incorporated to realize process control and coordination. Two indicators of places and transitions are employed to identify the key parts of HTCPN. A diagnostic analysis of group allocation is developed to determine the most appropriate way to deploy firefighters. A sensitivity analysis based on SHapley Additive exPlanations (SHAP) is conducted to identify the relevant importance of key factors on the performance of emergency. The proposed approach is applied at a Wuhan metro station. Results indicate that: (1) The firefighting and rescue part is of uppermost priority for emergency rescue time controlling; (2) Among all 19 influential factors, six key factors are tested as exerting more significant impacts on the performance of emergency; (3) The proposed model could significantly improve the efficiency of the emergency response, reducing the overall response time by 14.3%. The novelty of this research lies in the developed digital emergency rescue model, which is capable of taking uncertainties underlying complex coordination among various departments into account. The model has the potential to provide basic technology to support decision-making and help develop a human-centered emergency rescue system.
In chemical membrane cleaning, the challenge is to efficiently remove irreversible fouling while minimizing the impact on membrane materials. Particularly, traditional hypochlorite cleaning will ...further lead to the generation of toxic halogenated by-products. To address these issues, a combined system composed of peroxymonosulfate and chloride (PMS/Cl−) was applied to clean irreversible-humic-acid-fouled polyethersulfone (PES) membranes. After fouled membranes were soaked for 1 h in a PMS/Cl− solution (10 mM/15 mM) at 25 °C under neutral conditions, 94% flux recovery and 96% resistance removal were realized. Surface properties of virgin and cleaned membranes were very similar, confirming the effectiveness of the PMS/Cl− solution in removing irreversible foulants. The stability of membrane separation performance during multiple fouling and cleaning cycles further confirmed the minimal impact on membrane materials. Rapid diminution of the peaks centered in the region of fulvic-like and humic-like components, monitored under 3D-fluorescence for the cleaning solution, was attributed to PMS-catalyzed chlorination, thereby revealing the primary foulant detachment mechanism. Crucially, the approach exhibited lower toxicity than hypochlorite, as evidenced by reduced halogenated by-products and lower acute toxicity to Photobacterium phosphoreum T3. Overall, this novel cleaning system is promising for the efficient and environmentally friendly removal of irreversible organic foulants in practical water-treatment.
Display omitted
•The PMS/Cl− system was utilized to clean the fouled membrane for the first time.•The 94% flux recovery and the 96% resistance removal were achieved by PMS/Cl−.•PMS was verified as a catalyst for in-situ generated HClO in the cleaning process.•Membranes cleaned with PMS/Cl− exhibited enhanced stability and durability.•Lower levels of toxicity with PMS/Cl− compared to commercial method using HClO.
Purpose The paper aims to develop a useful tool that can reliably and accurately find the critical paths of high-rise buildings and provide optimal solutions considering the uncertainty based on ...Monte Carlo simulation (MCS) to enhance project implementation performance by assisting site workers and project managers in high-rise building engineering. Design/methodology/approach This research proposes an approach integrating the improved nondominated sorting genetic algorithm II (NSGA-II) considering uncertainty and delay scenarios simulated by MCS with the technique for order preference by similarity to an ideal solution. Findings The results demonstrate that the proposed approach is capable of generating optimal solutions, which can improve the construction performance of high-rise buildings and guide the implementation management for shortening building engineering project schedule and cost under the delay conditions. Research limitations/implications In this study, only the construction data of the two floors was focused due to the project at the construction stage, and future work can analyze the whole construction stage of the high-rise building to examine the performance of the approach, and the multi-objective optimization (MOO) only considered two factors as objectives, where more objectives, such as schedule, cost and quality, can be expanded in future. Practical implications The approach proposed in this research can be successfully applied to the construction process of high-rise buildings, which can be a guidance basis for optimizing the performance of high-rise building construction. Originality/value The innovations and advantages derived from the proposed approach underline its capability to handle project construction scheduling optimization (CSO) problems with different performance objectives under uncertainty and delay conditions.
Undersized fraction from aged municipal solid waste (UFAMSW), as a kind of soil-like material, has been proved effective in providing a large amount of organic matter and nutrients for soil and ...plants. The characteristics and effectiveness of heavy metal pollution removal in UFAMSW attracted tremendous research interest from scientists recently. In this study, the heavy metal removal efficiencies and bioavailability of washing on contaminated UFAMSW were evaluated with three washing reagents including ethylene diamine tetra acetic acid (EDTA), citric acid (CA), and humic acid (HA). The effects of chelating agent concentration, pH, and washing time on metal removal were investigated and response surface methodology (RSM) was employed to optimize the washing conditions. The results indicated that the removal efficiencies of Cu, Zn, and Mn could be 53.68%, 52.12%, and 30.63% by EDTA/HA washing and 42.36%, 39.67% and 28.49% by CA/HA washing, respectively. The European Community Bureau of Reference (BCR) sequential extraction was applied to analyze the fraction change of heavy metals in UFAMSW before and after washing, and it was found that chelating agent combined with HA could contribute to the removal of the exchangeable fraction. Physical and chemical properties of UFAMSW were improved to some extent after washing with mixed HA and chelating agent and could achieve the quality standard of landscape gardening soil. Accordingly, the mixture of HA and other chelating agents could be a promising washing process for preparation of landscape gardening soil using UFAMSW.
Implications: Our manuscript studies the removal of heavy metals from the contaminated undersized fraction from aged municipal solid waste (UFAMSW). UFAMSW, as a kind of soil-like material, has been proved effective in providing a large amount of organic matter and nutrients for soil and plants however often limited by heavy metal pollution. The UFAMSW used in this experiment was collected after the excavation and screening-sorting of aged refuse from Changshankou Domestic Waste Sanitary Landfill in Wuhan City, Hubei Province, Southern China. This study investigated the effects of EDTA, CA, HA, mixed EDTA/HA, and mixed CA/HA washing on heavy metal removal (Cu, Zn, and Mn), bioavailability of residual heavy metal and properties. The effects of chelating agent concentration, pH, and washing time on metal removal were investigated and then response surface methodology was employed to optimize the washing conditions. The results showed that washing by CA/HA and EDTA/HA, had a higher removal efficiency of heavy metals (Cu, Zn, and Mn) in UFAMSW compared to single HA. Meanwhile, HA has a higher removal for exchangeable fraction of heavy metals, the exchangeable concentration of Cu, Zn, and Mn in CA/HA and EDTA/HA washed UFAMSW were lower compared with UFAMSW washed by single CA and EDTA. Thus, mixing HA with EDTA or CA makes a less risk to environmental and the removal efficiency is acceptable. Additionally, CA/HA and EDTA/HA washing tend to improve soil physicochemical properties and soil fertility. Thus, mixing HA with different washing agent are potential methods for preparation of landscape gardening soil using UFAMSW.
Traditionally, the development of osteoarthritis (OA) is associated with factors such as aging and injure, but more and more epidemiological and biological evidence suggests that the disease is ...closely related to metabolic syndrome and metabolic components. Ubiquitin-specific protease 3(USP3), a member of the USPs family, is a specific protease capable of cleavage of ubiquitin chains linked by proline residues. In our presented study, we firstly found that USP3 expression level was decreased in OA. USP3 overexpression inhibited IL-1β induced chondrocytes apoptosis and nuclear factor κB (NF-κB) activation. USP3 knockdown induced chondrocytes apoptosis and activated NF-κB pathway. USP3 interacts with TRAF6 (tumor necrosis factor-receptor-associated factor 6), which is an essential adaptor protein for the NF-κB (nuclear factor κB) signaling pathway and plays important roles in inflammation and immune response. IL-1β treatment up-regulated the polyubiquitination of TRAF6 in chondrocytes, which was attenuated when USP3 was forced expression. Our study mechanistically links USP3 to TRAF6 in osteoarthritis development. Moreover, these data support the pursuit of USP3 and TRAF6 as potential targets for osteoarthritis therapies.
•USP3 expression level was decreased in OA.•USP3 inhibited IL-1β induced chondrocytes apoptosis and NF-κB activation.•USP3 interacted with TRAF6 and attenuated IL-1β-induced TRAF6 ubiquitination.
As one of the key technologies in production process scheduling, line balancing has played a decisive role in improving productivity and increasing utilization efficiency. In this work, a novel ...approach for balancing manual automobile assembly lines is proposed within an improved genetic algorithm framework, where task sequence chromosomes are coded with satisfaction of precedence constraints and partitioned dynamically under the restricts of unidirectional stations on assembly line. Furthermore, the initial populations, chromosome selection schemes for next generation, crossover and mutation operators are reformed subsequently and hence the assembly balancing problem can be solved to the optimality or near optimality. The computational studies and comparisons have proven the validity and feasibility of the proposed approach in automobile assembly production lines.