少样本文本分类的多任务原型网络

少样本文本分类中,原型网络对语义利用不足、可迁移特征挖掘不够,导致模型泛化能力不强,在新任务空间中分类性能不佳。从模型结构、编码网络、度量网络等角度提高模型泛化性,提出多任务原型网络(multiple-task prototypical network, MTPN)。结构上,基于原型网络度量任务增加辅助分类任务约束训练目标,提高了模型的语义特征抽取能力,利用多任务联合训练,获得与辅助任务更相关的语义表示。针对编码网络,提出LF-Transformer编码器,使用层级注意力融合底层通用编码信息,提升特征的可迁移性。度量网络使用基于BiGRU的类原型生成器,使类原型更具代表性,距离度量更加准确。实验表明,MTPN在少样本文本情感分类任务中取得了91.62%的准确率,比现有最佳模型提升了3.5%以上;在新领域的情感评论中,基于五条参考样本,模型对查询样本可获得超过90%的分类准确率。     

基于代理认证系统中CA的设计与实现

本文首先对移动警务中基于代理的身份认证系统的总体结构进行了介绍,然后依据X.509标准并结合移动警务系统的具体需求,提出了证书颁发机构的总体设计方案,包括了系统的总体结构、以及所使用证书的格式、编码、类型及存储的详细说明,最后,本文着重阐述了证书颁发机构所具备的各项功能。     

基于统一编码的信息孤岛集成技术研究

信息编码是信息孤岛集成的基础和关键技术。文章分析了信息集成环境对信息分类编码的要求,从信息编码多约束目标出发,采用AHP法分析了各约束目标对编码结构的影响权重,构建了一种基于面向对象技术的统一编码柔性结构模型,并对模型进行了描述和实例化,该模型最终实现了各类信息编码结构形式的统一。以统一编码结构模型为基础,提出了基于统一编码的信息孤岛集成技术实现方案,从基于统一编码的应用系统接口技术和基于统一编码的标准视图访问接口技术两个方面描述了集成方案的实现机理。     

联动油缸机构的遗传优化设计

为使喷浆机器人纵向进给机构与巷道中线始终保持平行 ,采用了一组油缸联动机构 ,并用改进的遗传算法对其进行了优化设计。改进后的遗传算法采用浮点数编码 ,用惩罚函数处理等式和不等式的约束问题 ,使之适用于所有的非线性约束优化问题     

工程布局CAD系统设计与实现

针对工程布局设计的复杂性,提出了开放式的工程布局设计CAD系统的结构框架,讨论了实现该系统的关键方法,包括自动化设计进程建模、基于功能特征的待布物和布局空间几何近似建模、布局知识和布局约束的表达、布局方法应用.基于该系统框架,实现了计算机辅助工程布局设计原型系统,并以车辆动力舱布局设计为背景,对原型系统的有效性进行了验证.     

一种编码约束的医学图像分割方法

深度学习的优势在于其具有深层次的特征提取结构,而随着层数的增加以及激活函数的影响往往会导致其编码能力下降.基于此本文提出了一种基于U-Net和FCN网络进行编码约束的方法,并应用到医学图像分割上.编码约束结构以U-Net和FCN全卷积网络模型架构为主体,对网络最后一层使用Sigmoid激活函数的1×1卷积层进行特征约束...     

基于网格计算的机构投资者IRMIS研究

在评述集成风险管理理论进展的基础上，指出了机构投资者集成风险管理信息系统(IRMIS)的复杂自适应系统(CAS)特征；引入网格计算思想提出了基于网格计算的机构投资者IRMIS的概念、特点及其系统原型，该原型具有多智能体(Multi-Agent)的良好结构，并基于管理熵对原型各Agent间的耦合关系进行了优化，强调风险管理战略、组织、方法、文化和过程诸要素的集成性。     

发现含有第一类项目约束的频繁集的快速算法

与Apriori-like类型的算法相比,Zaki提出的基于垂直数据库结构及基于网络理论的算法将关联规则挖掘的运行速度提高了一个数量级,并且这些算法非常适合挖掘低支持度、长模式的关联规则。以Ecalt算法为原型,讨论了如何将项目约束引入关联规则挖掘过程的问题,从理论上证明了引入约束后的Eclat＋算法可以大大提高算法的效率和速度,并对相关的算法进行了比较。     

H.264高清编码器系统的设计与仿真

常用软件编码方法在用于H.264标准时,很难实现对高清视频的实时解码。为此,提出软硬件协同设计的方法,对H.264高清编码器系统进行软硬件划分及任务分配,分别设计软件部分和硬件部分。建立软硬件协同仿真环境,对整个系统的功能和性能进行仿真,采用约束随机验证的方法提高验证的功能覆盖率。仿真结果表明,该编码器系统能够对 1 024×768像素的图像帧进行编码,满足实时编码的要求。     

CMG机构的优化编码程序开发

CMG机构是一种密码鉴别机构,齿牙编码是其核心设计问题。应用贪婪算法求解CMG机构优化编码的顶点着色问题,具有时间复杂度低、易于编程的优点,在大多数情况下可满足应用需求。基于这一思路,采用Visual Basic编写了一个用户界面友好的CMG机构优化编码程序。该程序包括两个功能定义清晰的模块,即编码模块和校验模块,可自动求解、校验优化编码。     

Automatic Virtual Entity Simulation of Conceptual Design Results—Part Ⅰ:Symbolic Scheme Generation and Identification

The development of new products of high quality,low unit cost,and short lead time to market are the key elements required for any enterprise to obtain a competitive advantage.For shorting the lead time to market and improving the creativity and performances of the product,a rule-based conceptual design approach and a methodology to simulate the conceptual design results generated in conceptual design process in automatical virtual entity form are presented in this paper.This part of paper presents a rule-based conceptual design method for generating creative conceptual design schemes of mechanisms based on Yan's kinematic chain regeneration creative design method.The design rules are adopted to describe the design requirements of the functional characteristics,the connection relationships and topological characteristics among mechanisms.Through the graphs-based reasoning process,the conceptual design space is expanded extremely,and the potential creative conceptual design results are then dug out.By refining the design rules,the solution exploration problem is avioded,and the tendentious conceptual design schemes are generated.Since mechanical,electrical and hydraulic subsystems can be transformed into general mechansims,the conceptual design method presented in this paper can also be applied in the conceptual design problem of complex mechatronic systems.And then the method to identify conceptual design schemes is given.     

Using sign algebra for qualitative spatial reasoning about the configuration of mechanisms

This paper presents a novel approach to qualitative spatial reasoning about the spatial configuration of mechanisms that could be applied in the earlier phases of mechanism design. The basic idea in this approach is to predict all possible spatial configurations of a mechanism using sign algebra. This paper addresses spatial configuration problems of direction, position, and orientation of mechanisms. Spatial configuration matrices are introduced to describe and manipulate the spatial information of primitive mechanisms. Using these matrices, the overall spatial configurations of a complicated mechanism are constructed and then they are expressed by spatial configuration state matrices. These matrices are composed based on sign algebra and used to reason about all possible spatial solutions of a mechanism. The qualitative spatial reasoning proposed in this paper could provide useful prior information on the spatial configuration of mechanisms earlier in the conceptual design of mechanisms. The proposed approach is illustrated with a design example.     

Automated functional design of engineering systems

This paper presents a prototype intelligent system, the knowledge-based functional design automation system (KBFDA) for automating the functional design process of engineering products/systems. An integrated knowledge representation scheme combines rule-based and object-oriented representation methods to represent functions and function related design characteristics in an intelligent design environment. A knowledge-based functional reasoning strategy uses this representation to automatically generate physical behaviors from desired functions or behaviors. The required behaviors are then combined in different configurations to develop a set of potential concept variants that meet the functional requirements and functional constraints given in a design specification. Finally, the variants are ranked according to the degree to which they meet non-functional constraints. The variant with the lowest rank (score) is selected as the best solution. A case study design of a terminal insertion unit is presented to demonstrate the practicality of the proposed approach.     

Conceptual design, manufacturability evaluation and preliminary process planning using function-form relationships in stamped metal parts

A large number of design decisions are made during the conceptual design of a part. However, there are few representation and reasoning tools for decision support during conceptual design. The conceptual design stage is characterized by a lack of complete geometric information. Existing geometric modelers require complete geometric information, while a functional reasoning methodology using a <verb, noun > representation is typically too terse. In this paper, we present a new representation called sketching abstraction for conceptual design, using the function-form relations in a design. The functionally critical part of the geometry is presented using a set of functional features, while the rest of the geometry is abstracted as a set of linkages. Part functionality is correlated with the sketching abstraction using data structures called function-form matrices. The sketching abstraction is annotated using a set of primitives, and a set of grammar rules are used to extract canonical relationships between the functional features. The sketching abstraction can be used for extracting designs that are geometrically dissimilar but functionally similar, thus providing the designer with ideas for design alternatives. The sketching abstraction can also be used to carry out domain-dependent manufacturability evaluation checks. A further use of sketching abstractions is to initiate the development of a process plan for manufacturing. Sketching abstractions are related to the solid model of a part. Thus, this representation provides a link between pure functional and pure geometric representations. The domain of application is stamped metal parts. We present the part functionality and the features used in this domain. We also illustrate the use of sketching abstractions for conceptual design, manufacturability evaluation and preliminary process planning.     

Case-Based Reasoning in Conceptual Design

The framework for supporting conceptual design using case-based reasoning is studied. Adesign knowledge representation scheme is presented. It is capable of representing both generalized and specific design knowledge. A two-phase case retrieval strategy, called index elaboration, is described which supports both requirement-function transformation and function-feature transformation at conceptual design stage. We propose a prototype system and discuss in detail the realization of supporting conceptual design by CBR.     

A Multidimensional and Multigranular Model of Time for Medical Knowledge-Based Systems

In temporal reasoning there are two interrelated issues; how to model time per se and how to model occurrences. In medical temporal reasoning the need for multiple granularities and multiple conceptual temporal contexts arises in relation to a model of time. Some occurrence can then be expressed with respect to different temporal contexts. This paper presents a multidimensional and multigranular model of time for knowledge-based problem solving, primarily for medical applications. Both the conceptual issues and the design issues underlying the implementation of the proposed model are discussed. The presented model of time has been developed in the context of a time ontology for medical knowledge engineering, whose principal primitives are the time-axis and the time-object. The notion of a time-axis constitutes the primitive for the proposed model of time, while the notion of a time-object aims to integrate time with other essential forms of knowledge, such as structural and causal knowledge, in the expression of different types of occurrences, thus resulting in the integral embodiment of time in such occurrences. The notion of a time-object and the overall ontology of occurrences is given only a cursory mention in this paper. The focus of the paper is the time model. More specifically, the paper presents the notion of a time-axis in the context of the overall time ontology and discusses at length the two classes of time-axes, namely the atomic axes and the spanning axes. The assertion language which has been developed, for the entire ontology, for the expression of axioms (deductive rules and integrity constraints), attribute constraints and propagation methods is presented and illustrated. The implementation of the time model in terms of a layered object-based system is also presented.     

面向对象知识模型的构建

介绍了面向对象技术设计知识库模式的方法。强调知识模型构建的重点是在表达、验证及推导设计限制条件，而不是传统意义上的数据。设计概念仍沿用类表示，分3个部分：决策参数、表现参数及限制条件。决策参数是从限制条件中抽象出来的数据，代表设计对象的属性或状态；表现参数是对决策参数的计算和度量；限制条件规范及限制表现参数的域值，可视为类不变量，每个类对象都必需满足之。设计概念之间的相互联系则用关联来阐述。传统面向对象技术考虑的是数据在关联中的横向、纵向传递，提出关联类以捕获数据，同时增加关联的限制条件作为关联不变量。在继承和整件-部件关联的基础上，增加了特性关联。模型构建构建以汽车制造中模板设计为例加以说明。     

Design candidate identification using neural network-based fuzzy reasoning

Conceptual design has profound impact on success of a product design. Identification of the best conceptual design candidate is a crucial step as design information is not complete and design knowledge is minimal at conceptual design stage. This paper presents a method for design candidate evaluation and identification using neural network-based fuzzy reasoning. The method consists of the following steps: (1) acquisition of customer needs and ranking of their importance, (2) establishment of measurable metrics and their relations with customer needs, (3) development of design specifications and initial evaluation of design candidates, and (4) evaluation and identification of design candidates based on design specifications and customer needs using neural network-based fuzzy reasoning. A case study is given to show the effectiveness of the proposed method and associated algorithms.     

A smart conflict resolution model using multi-layer knowledge graph for conceptual design

Reducing the impact of conflicts on requirement-function-structure mapping in the early stage of product design is an important measure to achieve conceptual innovation, which relies on accurate reasoning of multi-domain knowledge. As product requirements become more personalized and diverse, traditional discrete knowledge organization and reasoning methods are difficult to adapt to the challenges of continuity and precision in conceptual solution. Knowledge graphs with complex networks have obvious advantages in association detection, knowledge visualization, and explainable reasoning of implicit knowledge, which offer innovative opportunities for conflict resolution in conceptual design. Therefore, a smart conflict resolution model using a multi-layer Knowledge Graph for Conceptual Design(mKGCD) is proposed in this study. A knowledge expression form of FBS-oriented design patent vocabulary is proposed, which is used for knowledge entity recognition and relation extraction based on natural language processing. A label mapping method based on inventive principles is used for patent classification and a four-layer semantic network for conflict resolution is constructed. Through semantic distance calculation, the designer's requirements for function/behavior/structure are smart deployed to obtain appropriate knowledge. A case study of the conceptual design of a collapsible installation and handling equipment demonstrates the feasibility of the proposed approach. The proposed method can not only meet the functional solution and innovation in the context of different design requirements, but also effectively improve the design efficiency in the iterative design process by means of multiple meanings of one graph.     

基于产品动态模型的智能快速响应设计理论和方法研究

文中将产品动态模型应用于快速响应变型设计中,研究了快速响应变型设计智能化理论和方法,提出了采用规则推理和实例推理相结合的混合推理策略;并针对设计的过程性,将变型设计分为概念变型设计、装配变型设计和产品定义变型设计;最后分析了基于产品动态模型的智能变型设计过程。