排序支持的交互数据分类算法及其应用

交互分类是解决数据复杂分类问题的主要手段之一。在现有的大多交互分类系统中，用户能准确识别数据类别，但在有些分类场景中,类别之间的顺序关系更容易被识别，为此，提出一种排序支持的交互数据分类算法。为提升交互分类精度，引入数据的顺序信息，为降低标记难度，提出候选样本推荐策略。另外,提出一种评估分类算法性能的可视化方法，用包含基本车况、交通违法记录、交通事故记录等信息的车辆数据集进行实验验证，将相关车辆分为高危车辆、中危车辆、低危车辆3类，算法的分类结果模型一致度达近98%，验证了方法的有效性。     

互联网B/S架构交互式软件在核心概念教学中的应用*——以国际IB课程中“电解质”内容为例

基于核心概念教学的思路,国际IB课程互联网B/S架构交互式软件能够使学生围绕“电解质”核心概念,通过自行操作软件获得多个从微观到宏观、定性与定量相结合、可视化的实验证据,有利于学生深度理解概念与概念之间的逻辑和发展关系,促进学科知识系统框架的形成。     

A NEW CYCLIC J-INTEGRAL FOR LOW-CYCLE FATIGUE CRACK GROWTH

The constitutive equation under the low-cycle fatigue (LCF) was discussed, and a two-dimensional (2-D) model for simulating fatigue crack extension was put forward in order to propose a new cyclic J-integral. The definition, primary characteristics, physical interpretations and numerical evaluation of the new parameter were investigated in detail. Moreover, the new cyclic J-integral for LCF behaviors was validated by the compact tension (CT) specimens. Results show that the calculated values of the new parameter can correlate well with LCF crack growth rate, during constant-amplitude loading. In addition, the phenomenon of fatigue retardation was explained through the viewpoint of energy based on the concept of the new parameter.     

ORSIS – News and further developments

The paper deals with the simulation program Off Road Systems Interactive Simulation (ORSIS) which is the w,orldwide leading simulation tool for off road driving of wheeled vehicles. Even though the present state of development allows a very realistic simulation of the man–vehicle–terrain system, there is continuous further improvement in the computer program’s detail. Some of the recent innovations integrated into the program are presented in this paper. These examples were chosen in a way that the scope of the further developments can be demonstrated. They represent three main directions of the work carried out: the refining of the tire–soil-model, the integration of new technologies and subsystems into the vehicle model and the improvement of the man–machine–interface especially in driving simulators.The first part of the paper describes a further development of the tire–soil-model. A significant improvement has been achieved to include the influence of slippery surfaces on traction in combination with the tire tread pattern. Results from finite element method (FEM) as well as real measurements were used to build up an approach, which qualitatively allows the influence of the positive–negative portion of the tire tread and the lug height of the tire tread on traction to be considered. The basic idea is very simple and straightforward. Moreover the calculation costs are very low, so the enhancement does not affect real time operation.In the second part a physical model for the central tire inflation system (CTIS) is presented. With this model it is possible to simulate the complete pneumatic system of a CTIS, including the air compressor with an accumulator, the pressure line and the wheel valves. The components are modelled by their physical parameters, so an adaptation to different existing tire-pressure-control-systems (TPC) can be made. The paper presents a short review of the modelling and a first validation using real measurements. Furthermore the influence of each parameter, e.g. the discharge flow of the compressor on the inflation time, is presented.The third part of the paper describes a further development of the visualization system. The ORSIS OpenGL graphic engine was separated from the main ORSIS simulation and can be run on different PCs controlled via a network. It is therefore possible to build up very cheap multi-channel visualization systems using consumer PCs running under LinuX. The fact that ORSIS itself is running on a normal PC allows the assembly of comparatively cheap driving simulators of a high end simulation quality.     

动态J积分及Rice J积分一致性比较与分析

首先回顾了动态J积分的发展进程,然后给出了两种常用动态J积分和Rice J积分的数学计算式,并对它们的一致性加以讨论,由此提出一种广义J积分的定义。然后通过算例,基于ANSYS软件,建立平面应变情形下纯I型中心裂纹模型,并利用编写的APDL程序计算了各个J积分,通过结果比较,进一步验证了在动态加载条件下,动态J积分比Rice J积分具有更好的路径守恒性,但仍可以看出它们的取值具有一定的一致性。最后,通过J积分间接法和裂尖位移外推法分别计算裂纹动态应力强度因子,所得结果吻合很好,验证了编写程序的正确性。     

波纹钢屋盖结构板件相关屈曲的试验研究与分析

拱型波纹钢屋盖结构的受力机理异常复杂，波纹、加工制作过程中产生的变形和残余应力等初始缺陷在理论分析中很难精确定量描述。论文对波纹钢屋盖结构梯形截面板组进行了轴心和偏心压力作用下稳定承载能力的试验研究，并与理论结果进行了比较分析，试验与理论结果吻合较好。试验结果表明波纹板组具有较高的局部屈曲承载能力，但几乎无屈曲后承栽能力，与平板板组屈曲特性相比有显著差异。     

剪胶接结构的混合模式断裂分析

给出了Ｇｏｌａｎｄ－Ｒｅｉｓｓｅｎｅｒ型剪接结构的Ｊ积分表达在被粘物体小转角和胶层很薄的假设下与路径无关。Ｊ积分代表胶层厚度与胶层内最大能量密度的乘积。对单剪接接头，本文计算了裂纹在胶层内和界面上两种情况下的局部Ｉ型和Ⅱ型应力强度因子之比，即局部载荷相角，从而给出了单剪接接头的断裂判据。     

An extension of the J-integral to piezoelectric materials

This paper provides an analysis of the crack propagation criterion for a thin piezoelectric plate with a symmetry of order six. On the basis of Gol’denveizer’s asymptotic integration method or Destuynder’s unidirectional zoom technique, we obtain an extension of the purely mechanical J-integral to piezoelectric materials, with a dependence of the gradient of energy of the plate only on zeroth order terms of asymptotic expansions.     

New Interfaces and Approaches to Machine Learning When Classifying Gestures within Music

Interactive music uses wearable sensors (i.e., gestural interfaces—GIs) and biometric datasets to reinvent traditional human–computer interaction and enhance music composition. In recent years, machine learning (ML) has been important for the artform. This is because ML helps process complex biometric datasets from GIs when predicting musical actions (termed performance gestures). ML allows musicians to create novel interactions with digital media. Wekinator is a popular ML software amongst artists, allowing users to train models through demonstration. It is built on the Waikato Environment for Knowledge Analysis (WEKA) framework, which is used to build supervised predictive models. Previous research has used biometric data from GIs to train specific ML models. However, previous research does not inform optimum ML model choice, within music, or compare model performance. Wekinator offers several ML models. Thus, we used Wekinator and the Myo armband GI and study three performance gestures for piano practice to solve this problem. Using these, we trained all models in Wekinator and investigated their accuracy, how gesture representation affects model accuracy and if optimisation can arise. Results show that neural networks are the strongest continuous classifiers, mapping behaviour differs amongst continuous models, optimisation can occur and gesture representation disparately affects model mapping behaviour; impacting music practice.