478 research outputs found
Codification and research for mass timber buildings in selected seismic regions: An overview
No full textSince the late 2000's, many countries around the world have experienced rising interest in mass timber construction in commercial and mixed-use building applications. Globally, wood design and research communities have invested heavily in research and development (R&D) to make mass timber viable for large scale, multi-story buildings, targeting commercial markets. Due to the difference in historical design/fabrication practices, local regulatory rules, and cultural differences, the status of mass timber codification and research development are not uniform throughout the world. In this paper, an overview of recent trends in manufacture, codification, and research on mass timber systems is provided. Specifically, this overview was divided into five distinct topics, namely mass timber material standards, mandatory building/design requirements, non-mandatory design guidelines, different approaches to lateral design, and notable mass timber research efforts in the recent decades. Due to the limitation of the authors' experience, only selected regions around the world are covered in this review
A Reexamination of the Establishment of the Liangshuifa 兩稅法 with the Concept of liang chu yi zhi ru 量出以制入 of Yang Yan 楊炎
No full textTraditionally, the historical significance of the Liangshuifa 兩稅法 has been emphasized from the standpoint of economic and political history, as exemplified by the principle of liang chu zhi ru 量出制入 and the theory of yi fan zhen chao 抑藩振朝 of Hino Kaisaburo 日野開三郎. In the history of Chinese economic thought, it is said that it was Yang Yan 楊炎 who first proposed the concept of liang chu wei ru 量出為入 and the implementation of the Liangshuifa strengthened central finance and encouraged the reconstruction of the tax system that had not functioned well since the An-Shi Rebellion. This paper reexamines this perspective. Regarding the idea of liang chu yi zhi ru 量出以制入 in Yang's plan, previous research has misunderstood its original intention. In fact, it meant determining fiscal receipts according to people's ability to bear taxes, and it was the tax principle in the newly conceived fiscal income system. Regarding the issue of strengthening central finances, we reject it based on an examination of the history of the failure of the family registration in the 14th year of Dali 大曆, which has not previously been noted. The introduction of the policy that immobilized the entire amount of tax collected and abortion of the merger of the miscellaneous duties (雑徭) show that the central government compromised with local governments, which were reluctant to reform the tax system, and consequently conceded. Yang's intent to concentrate power in the central government was not realized. The actual implementation of taxation took the form of a balance between securing regional fiscal interests and the increase in central fiscal revenue, and did not weaken the financial power of the provinces and prefectures. The greatest concern of the central government lay in securing that the established tax revenue be delivered. The Liangshuifa deviated totally from the original intention. The establishment of the Liangshuifa does not indicate centralization, but instead means that the central government abandoned the fundamental means of national fiscal management that had existed since the establishment of the Legislative System, that is, the way the central government controlled local taxes through the family register
Determination of Platinum-Group Elements in Geological Samples by Isotope Dilution-Inductively Coupled Plasma-Mass Spectrometry Combined with Sulfide Fire Assay Preconcentration
No full textA method was developed for the determination of platinum-group elements (PGE) in geological samples by isotope dilution-inductively coupled plasma-mass spectrometry combined with sulfide fire assay preconcentration. Samples were fused and PGE analytes were concentrated in sulfide buttons. The buttons were dissolved using HCl leaving PGE analytes in insoluble residues, which were digested in HNO3 and simultaneously processed for the distillation of Os. The remaining solutions were further prepared for the purification of Ru, Rh, Pd, Ir and Pt using a tandem assembly of cation and Ln resin columns. The eluents were directly analysed by membrane desolvation-ICP-MS. Ruthenium, Pd, Os, Ir and Pt were determined by isotope dilution, whereas Rh was determined by conventional reference material calibration combined with Ir-193 as the internal standard element. The method was validated using a series of PGE reference materials, and the measurement data were consistent with the recommended and the literature values. The measurement precision was better than 10% RSD. The procedural blanks were 0.121ng for Ru, 0.204 for Rh, 0.960ng for Pd, 0.111ng for Os, 0.045ng for Ir and 0.661ng for Pt, and the limits of detection (3s) were 0.011ng g(-1) for Ru, 0.008ng g(-1) for Rh, 0.045ng g(-1) for Pd, 0.009ng g(-1) for Os, 0.006ng g(-1) for Ir and 0.016ng g(-1) for Pt when a test portion mass of 10g was used. This indicates that the proposed method can be used for the determination of trace amounts of PGE in geological samples
Exploiting a robust biopolymer network binder for an ultrahigh-areal-capacity Li–S battery
No full textHigh-loading electrodes play a crucial role in the practical applications of high-energy-density batteries,which are especially challenging for lithium–sulfur (Li–S) batteries. Herein, a mechanically robust network binder was constructed by weaving dual biopolymers (i.e., guar gum and xanthan gum) via the intermolecular binding effect of extensive functional groups in both polymers. This network binder was capable of effectively preventing polysulfides within the electrode from shuttling and, consequently, improved electrochemical performance. A remarkably high sulfur loading of 19.8 mg cm<sup>-2</sup> and an ultrahigh areal capacity of 26.4 mA h cm<sup>-2</sup> were achieved as a result of the robust mechanical properties of the network binder. This study paves a new way for obtaining high-energy-density batteries by the simple application of robust network biopolymer binders that are inherently low-cost and environmentally friendly
Distributed SLIDE: Enabling Training Large Neural Networks on Low Bandwidth and Simple CPU-Clusters via Model Parallelism and Sparsity
No full textMore than 70% of cloud computing is paid for but sits idle. A large fraction of these idle compute are cheap CPUs with few cores that are not utilized during the less busy hours. This paper aims to enable those CPU cycles to train heavyweight AI models. Our goal is against mainstream frameworks, which focus on leveraging expensive specialized ultra-high bandwidth interconnect to address the communication bottleneck in distributed neural network training. This paper presents a distributed model-parallel training framework that enables training large neural networks on small CPU clusters with low Internet bandwidth. We build upon the adaptive sparse training framework introduced by the SLIDE algorithm. By carefully deploying sparsity over distributed nodes, we demonstrate several orders of magnitude faster model parallel training than Horovod, the main engine behind most commercial software. We show that with reduced communication, due to sparsity, we can train close to a billion parameter model on simple 4-16 core CPU nodes connected by basic low bandwidth interconnect. Moreover, the training time is at par with some of the best hardware accelerators
High-iodine-loading quasi-solid-state zinc-iodine batteries enabled by a continuous ion-transport network
No full textZinc–iodine (Zn–I₂) batteries are promising candidates for next-generation large-scale energy storage systems due to their inherent safety, environmental sustainability, and potential cost-effectiveness compared to lithium-ion batteries. Their applications, however, have been limited by the sluggish Zn²⁺ transfer kinetics, severe polyiodide shuttling, and relatively low mass loading of iodine cathodes. Herein, we report a design strategy for a quasi-solid-state Zn–I₂ battery with a continuous 3D ion-transport network by integrating a thick iodine cathode and a bacterial cellulose hydrogel electrolyte. The polar bacterial cellulose fibers formed an interconnected network that provided abundant ion pathways for inward Zn²⁺ transport and also limited iodine species dissolution. The continuous 3D ion-transport networks were formed throughout the entire thick iodine cathode, resulting in a 10-times higher Zn-ion conductivity compared with the conventional-structured cathode. The quasi-solid-state Zn–I₂ battery based on the Zn anode and an integrated cathode delivered a reversible capacity of 176.6 mA h g⁻¹ and achieved long-term cycling for 900 cycles at 1C under an iodine loading of 20.0 mg cm⁻². The iodine loading can be further increased to 39.3 mg cm⁻² by adjusting the thickness of the cathode. Under a practical condition of low negative/positive ratio (N/P) of 2.1, an energy density of 56.4 Wh kg⁻¹ is achieved. This integrated electrode design provides guidelines for fabricating high-energy quasi-solid-state Zn ion batteries.Xin Yang, Minghao Xie, Zhijie Yan, Hang Ruan, Chunpeng Yang, Zaiping Guo, and Zi-Jian Zhen
Efficient Neural Network Architecture Search
No full textOne-Shot Neural Architecture Search (NAS) is a promising method to significantly reduce search time without any separate training. It can be treated as a Network Compression problem on the architecture parameters from an overparameterized network. However, there are two issues associated with most one-shot NAS methods. First, dependencies between a node and its predecessors and successors are often disregarded which result in improper treatment over zero operations. Second, architecture parameters pruning based on their magnitude is questionable. In this thesis, classic Bayesian learning approach is applied to alleviate these two issues. Unlike other NAS methods, we train the over-parameterized network for only one epoch before update network architecture. Impressively, this enabled us to find the optimal architecture in both proxy and proxyless tasks on CIFAR-10 within only 0.2 GPU days using a single GPU. As a byproduct, our approach can be transferred directly to convolutional neural networks compression by enforcing structural sparsity that is able to achieve extremely sparse networks without accuracy deterioration.Mechanical Engineering | Vehicle Engineerin
A Framework For Concept Drifting P2P Traffic Identification
No full textIdentification of network traffic using port-based or payload-based analysis is becoming increasing difficult with many Peer-to-Peer (P2P) application using dynamic ports, masquerading techniques, and encryption to avoid detection. To overcome this problem, several machine learning technique were proposed to classify P2P traffics. But in the real P2P network environment, new communities of peers often attend and old communities of peers often leave. It requires the identification methods to be capable of coping with concept drift, and updating the model incrementally. In this paper, we present a concept-adapting algorithm CluMC which is based on streaming data mining techniques to identify P2P applications in Internet traffic. The CluMC use micro-cluster structures which contain potential micro-cluster structures and outlier micro-cluster structures to classify the P2P traffic and discover the concept drift with limited memory. Our performance study over a number of real data sets that we captured at a main gateway router demonstrates the effectiveness and efficiency of our method. DOI: http://dx.doi.org/10.11591/telkomnika.v11i8.303
- …
