A method of constructing an inspiration library driven by user-perceived preference evaluation data for biologically inspired design

Biologically inspired design (BID) is one of the common methods for product design. To solve the problem of inaccurate acquisition of inspirational creatures due to the lack of user perception preference analysis, a data-driven intelligent service model for BID considering user perception needs is proposed based on Kansei engineering. Firstly, by extracting the perceptual features of creatures from the semantic source elements of products through mapping and encodes them, we proposed a data acquisition method based on intuitionistic fuzzy sets considering different customer preference distributions, bridging the gap caused by the asymmetry between designers and users. Secondly, the functional relationship between biometric features and user-perceived attributes is identified and predicted, and a predictive model of biodata considering user preferences is obtained by multiple linear regression analysis. Finally, based on the data clustering and reorganization theory to understand the organization and dynamics of the database, the construction of a BID library was completed, and the design resources in the library were used as analyzed knowledge for designers to plan design activities. Taking the bionic design of a UAV product as an example, a prototype of a computer-aided design service system was developed based on the theory proposed in the article, and the analyzed knowledge was used to improve the efficiency and science of the design, effectively verifying the usefulness of this study for design. To a certain extent, this study addresses the problem of cognitive limitations of designers and cognitive differences between designers and users, promotes the application of bioinspiration in product design, and improves the marketability of design solutions.     

Large domain-wall currents in epitaxial Au/BiFeO3/SrRuO3 thin-film capacitors with modulated oxygen vacancy and wall densities

Large domain wall (DW) conductivity in an insulating ferroelectric plays an important role in the future nanosensors and nonvolatile memories. However, the wall current was usually too small to drive high-speed memory circuits and other agile nanodevices requiring high output-powers. Here, a large domain-wall current of 67.8 μA in a high on/off ratio of ~4460 was observed in an epitaxial Au/BiFeO₃/SrRuO₃ thin-film capacitor with the minimized oxygen vacancy concentration. The studies from read current-write voltage hysteresis loops and piezo-response force microscope images consistently showed remaining of partially unswitched domains after application of an opposite poling voltage that increased domain wall density and wall current greatly. A theoretical model was proposed to explain the large wall current. According to this model, the domain reversal occurs with the appearance of head-to-head and tail-to-tail 180° domain walls (DWs), resulting in the formation of highly conductive wall paths. As the applied voltage increased, the domain-wall number increased to enhance the on-state current, in agreement with the measurements of current-voltage curves. This work paves a way to modulate DW currents within epitaxial Au/BiFeO₃/SrRuO₃ thin-film capacitors through the optimization of both oxygen vacancy and domain wall densities to achieve large output powers of modern domain-wall nanodevices.     

Influence of domain switching process on the electrical fatigue behavior of ferroelectrics

We report an experimental approach, designed based on the recent findings that domain switching in ferroelectric ceramics can be separated into three regimes during antiparallel electric field loading, to investigate the influence of domain switching process on the electrical fatigue behavior of ferroelectrics. Uniaxial compressive stress (−2 MPã -100 MPa) and thermal loading (20 °C–150 °C) were used to tune the domain switching process. Under the same loading condition, the bipolar electrical fatigue behavior of soft lead zirconate titanate ceramics was systematically characterized. The amplitude and frequency of the applied electric field are 2 kV/mm and 10 Hz, respectively. By analyzing the evolution of the domain switching process, combined with the measured polarization and strain response, as well as the cracks observed on the surface of the specimen, it is found that the fatigue of ferroelectric ceramics was mainly related to the domain switching process near the coercive electric field: the regime 2 defined in this paper. The underlying mechanism was further discussed by considering the interplay between the domain switching process with the main factors affecting the electrical fatigue of ferroelectrics, namely defect redistribution, charge carrier injection, and crack initiation.     

基于桥连接的词典学习方法的语义解析

现阶段的语义解析方法大部分都基于组合语义,这类方法的核心就是词典。词典是词汇的集合,词汇定义了自然语言句子中词语到知识库本体中谓词的映射。语义解析一直面临着词典中词汇覆盖度不够的问题。针对此问题,该文在现有工作的基础上,提出了基于桥连接的词典学习方法,该方法能够在训练中自动引入新的词汇并加以学习,为了进一步提高新学习到的词汇的准确度,该文设计了新的词语—二元谓词的特征模板,并使用基于投票机制的核心词典获取方法。该文在两个公开数据集(WebQuestions和Free917)上进行了对比实验,实验结果表明,该文方法能够学习到新的词汇,提高词汇的覆盖度,进而提升语义解析系统的性能,特别是召回率。     

两阶段领域自适应学习

针对领域自适应问题中源域和目标域的联合分布差异最小化问题,提出两阶段领域自适应学习方法.在第一阶段考虑样本标签和数据结构的判别信息,通过学习一个共享投影变换,使投影后的共享空间中边缘分布的差异最小.第二阶段利用源域标记数据和目标域非标记数据学习一个带结构风险的自适应分类器,不仅能最小化源域和目标域条件分布差异,还能进一步保持源域和目标域边缘分布的流形一致性.在3个基准数据集上的实验表明,文中方法在平均分类准确率和Kappa系数两项评价指标上均表现较优.     

An overlapping domain decomposition dual reciprocity method

We present a domain decomposition boundary integral equation method for two-dimensional partial differential equations. The overlapping Schwarz method is employed to improve the dual reciprocity boundary element method. The resulting algorithm turns out to be more accurate than a non-overlapping approach previously proposed. Some numerical results showing the improved accuracy and efficiency of the methods are given.     

带类型λ—演算

一、引言补演算是一个用函数描述计算的形式理论体系[1J,它由许多计算模型所组成。文〔1〕指出,应用认-演算比纯粹补演算更接近程序语言.但是由于内建函子和常数的引入常会发生一些问题,例如.常数O和函子+在机器中同样看待     

椭圆型偏微分方程的加法Schwarz方法

1.引 言 古典加法Schwarz方法(ASM)对于一般问题收敛很慢,在大多数情况下, ASM只能作为预条件子.另一方面,ASM的并行性能非常好,尤其适合大规模粗粒度并行计算,近年来随着并行机系统及并行计算的兴起,ASM重新受到重视.许多学者研究了怎样提高ASM的收敛速度[1,2,4].他们发现加法Schwarz方法之所以收敛慢是由于在内边界上采用     

A Refinement Calculus for Shared-Variable Parallel and Distributed Programming

Parallel computers have not yet had the expected impact on mainstream computing. Parallelism adds a level of complexity to the programming task that makes it very error-prone. Moreover, a large variety of very different parallel architectures exists. Porting an implementation from one machine to another may require substantial changes. This paper addresses some of these problems by developing a formal basis for the design of parallel programs in the form of a refinement calculus. The calculus allows the stepwise formal derivation of an abstract, low-level implementation from a trusted, high-level specification. The calculus thus helps structuring and documenting the development process. Portability is increased, because the introduction of a machine-dependent feature can be located in the refinement tree. Development efforts above this point in the tree are independent of that feature and are thus reusable. Moreover, the discovery of new, possibly more efficient solutions is facilitated. Last but not least, programs are correct by construction, which obviates the need for difficult debugging. Our programming/specification notation supports fair parallelism, shared-variable and message-passing concurrency, local variables and channels. The calculus rests on a compositional trace semantics that treats shared-variable and message-passing concurrency uniformly. The refinement relation combines a context-sensitive notion of trace inclusion and assumption-commitment reasoning to achieve compositionality. The calculus straddles both concurrency paradigms, that is, a shared-variable program can be refined into a distributed, message-passing program and vice versa. Received July 2001 / Accepted in revised form May 2002     

A process oriented semantics of the PRAM-language FORK

The parallel language FORK [1], based on a scalable shared memory model, is a PASCAL-like language with some additional parallel constructs. A PRAM (Parallel Random Access Machine) algorithm can be expressed on a high level of abstraction as a FORK program which is translated into efficient PRAM code guaranteeing theoretically predicted runtimes.

In this paper, we concentrate on those features of the language FORK related to parallelism, such as the group concept, a shared memory access and synchronous or asynchronous execution. We present a trace-based denotational interleaving semantics where processes describe synchronous computations. Processes are created or deleted dynamically and run asynchronously. Interleaving rules reflect the underlying CRCW (concurrent-read-concurrent-write) PRAM model.