42 research outputs found
Combined MPC and reinforcement learning for traffic signal control in urban traffic networks
No full textIn general, the performance of model-based controllers cannot be guaranteed under model uncertainties or disturbances, while learning-based controllers require an extensively sufficient training process to perform well. These issues especially hold for large-scale nonlinear systems such as urban traffic networks. In this paper, a new framework is proposed by combining model predictive control (MPC) and reinforcement learning (RL) to provide desired performance for urban traffic networks even during the learning process, despite model uncertainties and disturbances. MPC and RL complement each other very well, since MPC provides a sub-optimal and constraint-satisfying control input while RL provides adaptive control laws and can handle uncertainties and disturbances. The resulting combined framework is applied for traffic signal control (TSC) of an urban traffic network. A case study is carried out to compare the performance of the proposed framework and other baseline controllers. Results show that the proposed combined framework outperforms conventional control methods under system uncertainties, in terms of reducing traffic congestion. Green Open Access added to TU Delft Institutional Repository ‘You share, we take care!’ – Taverne project https://www.openaccess.nl/en/you-share-we-take-care Otherwise as indicated in the copyright section: the publisher is the copyright holder of this work and the author uses the Dutch legislation to make this work public.Team Bart De SchutterControl & SimulationDelft Center for Systems and Contro
A Hardware/Software Co-designed Partitioning Algorithm of Sparse Matrix Vector Multiplication into Multiple Independent Streams for Parallel Processing
No full textThe trend of computing faster and more efficiently has been a driver for the computing industry since its beginning. However, it is increasingly difficult to continue this trend because current CMOS technology cannot be down-scaled anymore due to physical restrictions. Consequently, to obtain the next major performance improvement, the focus is shifting from a technology-only optimization effort towards a system-level hardware-software co-design optimization strategy. In recent years, the move to heterogeneous computing has gained enormous traction with all the big names such as Intel, IBM, and NVIDIA investing heavily in this approach. This paradigm shift is characterized by traditional general-purpose processors offloading data to hardware accelerators, which are capable of exploiting parallelism to a significantly higher degree. An accelerator which has existed for decades but has recently risen to greater prominence is the field-programmable gate array (FPGA). The scientific computing community is also experiencing the need for higher computational power as their problem sizes increase. FPGAs make a promising candidate for their ability to tailor complex algorithms to specialized hardware circuits. A key algorithm to accelerate in this domain is the Sparse Matrix Vector Multiplication (SpMV). There do not exist many HLS (High-Level Synthesis) designs for this kernel, and the one designed using Vivado HLS exhibits significantly lower performance than the state-of-the-art. We argue that the most effective way to achieve speedup is by implementing multiple parallel pipelines so that multiple result values are produced in each cycle. Consequently, we develop an implementation agnostic partitioning algorithm for SpMV that splits the problem into independent streams. The HLS kernel performs well as a standalone unit, offering a speedup of up to 150x compared to the ARM coprocessor on the ZYNQ system and up to 4.6x to state-of-the-art Vivado HLS-based solutions. Our estimations show that the solution scales with an increasing number of resources.Computer Engineerin
Reply to comment on ‘Lost in translation: topological singularities in group field theory’
No full textIn Smerlak (2011 Class. Quantum. Grav.
28 178001) the author disputes the conclusion of our paper (Gurau 2010 Class. Quantum Grav.
27 235023). He claims that the Feynman graphs of three-dimensional group field theory always represent pseudo-manifolds. However,
Smerlak (2011) uses a different definition for pseudo-manifolds.
In order to apply the new definition Smerlak (2011) proposes a construction which cannot be implemented in a path integral by Feynman rules.
These two points invalidate the claims of Smerlak (2011).</jats:p
Innovation and Public Reform
No full textAbstractInnovation involves taking a risk and accepting a potential of failure. Thus, reporting to our public administrative system, the process of innovation is not normally associated with thinking and acting within the public sector. This is a real challenge, and if this challenge is overcome, innovation has the proper internal channels to take effect. Innovation is a dynamic process that changes the overall architecture of government, identify issues, challenges, develop new processes, creative, and selection and implementation of new solutions. Thus, in general the features of innovation coincides largely with the reform process. Important factor on promoting innovation is promoting a creative mindset. The innovation process is essential to increase public sector efficiency and for delivering quality and competitive public services. The purpose of this paper is to present and analyze innovation as a social phenomenon and how it is tangential to the public administration reform and identify the defining characteristics that classify the reform process as one innovative. Based on a review and analysis of the literature, acceptions of the reform process and innovation in the public sector, its objectives, motivational factors, barriers, and indissoluble link between reform and innovation will be presented in this paper
Triple dot system coupled to topological insulators: A numerical renormalization group analysis of the SU(3) attractive Anderson model
No full textStable passivation of cut edges in encapsulated n-type silicon solar cells using Nafion polymer
No full textIn this study, the edge passivation effectiveness and long-term stability of Nafion polymer in n-type interdigitated back contact (IBC) solar cells are investigated. For new module technologies such as half-cut, triple-cut, or shingled modules, cutting of the cells introduces unpassivated edges with a high recombination rate and this limits the module power. These cut edges can be “repassivated” after cutting and in this work Nafion polymer is used to achieve this. First, different edge types, namely emitter edges (n+/n/p+) and back surface field (BSF) edges (n+/n/n+), as well as different cutting techniques such as laser cut and cleave (L&C), thermal laser separation (TLS), and mechanical cleaving are evaluated. It is found that TLS and mechanical cleaving enable good repassivation on both BSF and emitter edges. Second, industrial-size IBC solar cells are made to assess the effect of the edge repassivation on performance. On 1/4-cut M2 size IBC cells with two emitter edges, efficiency is improved by over 0.3%abs. However, an efficiency improvement was not observed for similar cells with BSF edges, due to an insufficient passivation at the bulk edges. Last, the real-world stability of the Nafion repassivation is evaluated in industrially relevant module stacks by laminating the repassivated wafers with ethylvinylacetate (EVA) or polyolefin elastomer (POE) encapsulants and then exposing them to industry standard testing of 1000 h under damp heat conditions (85 °C, 85% relative humidity). The tests reveal that the repassivation is stable in EVA encapsulants but not in POE.Green Open Access added to TU Delft Institutional Repository 'You share, we take care!' - Taverne project https://www.openaccess.nl/en/you-share-we-take-care Otherwise as indicated in the copyright section: the publisher is the copyright holder of this work and the author uses the Dutch legislation to make this work public.Photovoltaic Materials and DevicesElectrical Sustainable Energ
The Influence of the Social, Political and Economic Impact on Human Resources, as a Determinant Factor of Sustainable Development
No full textAbstractRomanian society has seen an unpredictable political, economic and social dynamic due to political changes that took place in the late 90's, following until today a path dotted by numerous disturbing elements. Superimposed on European and global reformation, these changes have requested inclusion of Romanian policies into economic and social paradigms, which has caused major changes in all public administration systems. The HR segment, sensitive to the challenges of the external environment, perceived as a source of opportunity, but equally as a source of threat - by continuously changing laws, economic conditions, demographic policies, social and technological changes, etc. - strongly feels such changes. Thus, the major role of HR departments is a sound and efficient management of their resources, by adopting flexible adaptation strategies, based on accessing opportunities and avoiding threats. However, HR structures must consider staff response to consequences caused by these changes, in addition to external positive or negative impact. A major goal of sustainable development was the implementation of policies intended to optimize administrative capacity and professionalization of public services, as a measure of effective governance and good public policies. In achieving this, human resource occupies an important place, and the major challenge was to harmonize this segment, with influence in any governmental institution, with the effervescent environment to which it belongs. The study presented in this paper is based on a focus-group research that, by applying an interview matrix to the samples selected, analyzing and interpreting the responses, aims to highlight the political, social and economic impact on human resources, and the ability of organizations to adapt to these changes
An Overview of the Impact of Bacterial Infections and the Associated Mortality Predictors in Patients with COVID-19 Admitted to a Tertiary Center from Eastern Europe
No full text1. Background: Literature data on bacterial infections and their impact on the mortality rates of COVID-19 patients from Romania are scarce, while worldwide reports are contrasting. 2. Materials and Methods: We conducted a unicentric retrospective observational study that included 280 patients with SARS-CoV-2 infection, on whom we performed various microbiological determinations. Based on the administration or not of the antibiotic treatment, we divided the patients into two groups. First, we sought to investigate the rates and predictors of bacterial infections, the causative microbial strains, and the prescribed antibiotic treatment. Secondly, the study aimed to identify the risk factors associated with in-hospital death and evaluate the biomarkers’ performance for predicting short-term mortality. 3. Results: Bacterial co-infections or secondary infections were confirmed in 23 (8.2%) patients. Acinetobacter baumannii was the pathogen responsible for most of the confirmed bacterial infections. Almost three quarters of the patients (72.8%) received empiric antibiotic therapy. Multivariate logistic regression has shown leukocytosis and intensive care unit admission as risk factors for bacterial infections and C-reactive protein, together with the length of hospital stay, as mortality predictors. The ROC curves revealed an acceptable performance for the erythrocyte sedimentation rate (AUC: 0.781), and C-reactive protein (AUC: 0.797), but a poor performance for fibrinogen (AUC: 0.664) in predicting fatal events. 4. Conclusions: This study highlighted the somewhat paradoxical association of a low rate of confirmed infections with a high rate of empiric antibiotic therapy. A thorough assessment of the risk factors for bacterial infections, in addition to the acknowledgment of various mortality predictors, is crucial for identifying high-risk patients, thus allowing a timely therapeutic intervention, with a direct impact on improving patients’ prognosis
