A review of flexible halide perovskite solar cells towards scalable manufacturing and environmental sustainability

The perovskite material has many superb qualities which allow for its remarkable success as solar cells;flexibility is an emerging field for this technology.To encourage commercialization of flexible perovskite solar cells,two main areas are of focus:mitigation of stability issues and adaptation of production to flexible substrates.An in-depth report on stability concerns and solutions follows with a focus on Ruddlesden-Popper perovskites.Roll to roll processing of devices is desired to further reduce costs,so a review of flexible devices and their production methods follows as well.The final focus is on the sustainability of perovskite solar cell devices where recycling methods and holistic environmental impacts of devices are done.

Accelerating hybrid and compact neural networks targeting perception and control domains with coarse-grained dataflow reconfiguration

Driven by continuous scaling of nanoscale semiconductor technologies,the past years have witnessed the progressive advancement of machine learning techniques and applications.Recently,dedicated machine learning accelerators,especially for neural networks,have attracted the research interests of computer architects and VLSI designers.State-of-the-art accelerators increase performance by deploying a huge amount of processing elements,however still face the issue of degraded resource utilization across hybrid and non-standard algorithmic kernels.In this work,we exploit the properties of important neural network kernels for both perception and control to propose a reconfigurable dataflow processor,which adjusts the patterns of data flowing,functionalities of processing elements and on-chip storages according to network kernels.In contrast to stateof-the-art fine-grained data flowing techniques,the proposed coarse-grained dataflow reconfiguration approach enables extensive sharing of computing and storage resources.Three hybrid networks for MobileNet,deep reinforcement learning and sequence classification are constructed and analyzed with customized instruction sets and toolchain.A test chip has been designed and fabricated under UMC 65 nm CMOS technology,with the measured power consumption of 7.51 mW under 100 MHz frequency on a die size of 1.8×1.8 mm^2.

A Hadoop Performance Prediction Model Based on Random Forest

MapReduce is a programming model for processing large data sets, and Hadoop is the most popular open-source implementation of MapReduce. To achieve high performance, up to 190 Hadoop configuration parameters must be manually tunned. This is not only time-consuming but also error-pron. In this paper, we propose a new performance model based on random forest, a recently devel- oped machine-learning algorithm. The model, called RFMS, is used to predict the performance of a Hadoop system according to the system＇ s configuration parameters. RFMS is created from 2000 distinct fine-grained performance observations with different Hadoop configurations. We test RFMS against the measured performance of representative workloads from the Hadoop Micro-benchmark suite. The results show that the prediction accuracy of RFMS achieves 95% on average and up to 99%. This new, highly accurate prediction model can be used to automatically optimize the performance of Hadoop systems.

Big Data: Where Dreams Take Flight

From academia to industry, big data has become a buzzword in information technology. The US Federal Government is paying much attention to the big-data revolution. In 2012, fourteen US government departments allocated funds to 87 big-data projects [1].

Resource scheduling techniques in cloud from a view of coordination:a holistic survey

Nowadays,the management of resource contention in shared cloud remains a pending problem.The evolution and deployment of new application paradigms(e.g.,deep learning training and microservices)and custom hardware(e.g.,graphics processing unit(GPU)and tensor processing unit(TPU))have posed new challenges in resource management system design.Current solutions tend to trade cluster efficiency for guaranteed application performance,e.g.,resource over-allocation,leaving a lot of resources underutilized.Overcoming this dilemma is not easy,because different components across the software stack are involved.Nevertheless,massive efforts have been devoted to seeking effective performance isolation and highly efficient resource scheduling.The goal of this paper is to systematically cover related aspects to deliver the techniques from the coordination perspective,and to identify the corresponding trends they indicate.Briefly,four topics are involved.First,isolation mechanisms deployed at different levels(micro-architecture,system,and virtualization levels)are reviewed,including GPU multitasking methods.Second,resource scheduling techniques within an individual machine and at the cluster level are investigated,respectively.Particularly,GPU scheduling for deep learning applications is described in detail.Third,adaptive resource management including the latest microservice-related research is thoroughly explored.Finally,future research directions are discussed in the light of advanced work.We hope that this review paper will help researchers establish a global view of the landscape of resource management techniques in shared cloud,and see technology trends more clearly.

Graphdiyne oxide doping for aggregation control of hole-transportnanolayer in inverted perovskite solar cells

The strong aggregation tendency of hole transport material poly[3-(4-carboxylbutyl)thiophene-K(P3CT-K)restricts its further application in inverted perovskite solar cells(PSCs).Here,we report an effective strategy to address this issue and achieve the superior performance of inverted methylammonium lead triiodide(MAPbI3)PSCs,in which graphdiyne oxide(GDYO)doped P3CT-K nanocomposites are applied as the hole transport nanolayer(HTL).It is revealed that the strongπ–πstacking interaction occurs between GDYO and P3CT-K,which is proved by the blue shift of the absorption peak of P3CT-K nanolayer.The aggregation control via GDYO contributes to the property improvement of P3CT-K HTL.Moreover,the homogeneous coverage induces the growth of perovskite grain with larger size than that based on the undoped one.As a result,the optimized surface morphology,enhanced conductivity,charge extraction as well as better crystal quality,finally improve the device performance.An optimal power conversion efficiency of 19.06%is achieved,with simultaneously improved fill factor and short circuit current density.This work presents the potential of functional graphdiyne(GDY)in the development of highly efficient photovoltaic device.

Sensitivity analysis to reduce duplicated features in ANN training for district heat demand prediction

Artificial neural network(ANN)has become an important method to model the nonlinear relationships between weather conditions,building characteristics and its heat demand.Due to the large amount of training data re-quired for ANN training,data reduction and feature selection are important to simplify the training.However,in building heat demand prediction,many weather-related input variables contain duplicated features.This paper develops a sensitivity analysis approach to analyse the correlation between input variables and to detect the variables that have high importance but contain duplicated features.The proposed approach is validated in a case study that predicts the heat demand of a district heating network containing tens of buildings at a university campus.The results show that the proposed approach detected and removed several unnecessary input variables and helped the ANN model to reduce approximately 20%training time compared with the traditional methods while maintaining the prediction accuracy.It indicates that the approach can be applied for analysing large num-ber of input variables to help improving the training efficiency of ANN in district heat demand prediction and other applications.

Design and experimental validation of looped-tube thermoacoustic engine

The aim of this paper is to present the design and experimental validation process for a thermoacoustic looped-tube engine.The design procedure consists of numerical modelling of the system using DELTA EC tool,Design Environment for Low-amplitude ThermoAcoustic Energy Conversion,in particular the effects of mean pressure and regenerator configuration on the pressure amplitude and acoustic power generated.This is followed by the construction of a practical engine system equipped with a ceramic regenerator-a substrate used in automotive catalytic converters with fine square channels.The preliminary testing results are obtained and compared with the simulations in detail.The measurement results agree very well on the qualitative level and are reasonably close in the quantitative sense.

Prediction of office building electricity demand using artificial neural network by splitting the time horizon for different occupancy rates

Due to the impact of occupants’activities in buildings,the relationship between electricity demand and ambient temperature will show different trends in the long-term and short-term,which show seasonal variation and hourly variation,respectively.This makes it difficult for conventional data fitting methods to accurately predict the long-term and short-term power demand of buildings at the same time.In order to solve this problem,this paper proposes two approaches for fitting and predicting the electricity demand of office buildings.The first proposed approach splits the electricity demand data into fixed time periods,containing working hours and non-working hours,to reduce the impact of occupants’activities.After finding the most sensitive weather variable to non-working hour electricity demand,the building baseload and occupant activities can be predicted separately.The second proposed approach uses the artificial neural network(ANN)and fuzzy logic techniques to fit the building baseload,peak load,and occupancy rate with multi-variables of weather variables.In this approach,the power demand data is split into a narrower time range as no-occupancy hours,full-occupancy hours,and fuzzy hours between them,in which the occupancy rate is varying depending on the time and weather variables.The proposed approaches are verified by the real data from the University of Glasgow as a case study.The simulation results show that,compared with the traditional ANN method,both proposed approaches have less root-mean-square-error(RMSE)in predicting electricity demand.In addition,the proposed working and non-working hour based regression approach reduces the average RMSE by 35%,while the ANN with fuzzy hours based approach reduces the average RMSE by 42%,comparing with the traditional power demand prediction method.In addition,the second proposed approach can provide more information for building energy management,including the predicted baseload,peak load,and occupancy rate,without requiring additional building parameters.