Improving human robot collaboration through Force/Torque based learning for object manipulation

Human–Robot Collaboration (HRC) is a term used to describe tasks in which robots and humans work together to achieve a goal. Unlike traditional industrial robots, collaborative robots need to be adaptive; able to alter their approach to better suit the situation and the needs of the human partner. As traditional programming techniques can struggle with the complexity required, an emerging approach is to learn a skill by observing human demonstration and imitating the motions; commonly known as Learning from Demonstration (LfD). In this work, we present a LfD methodology that combines an ensemble machine learning algorithm (i.e. Random Forest (RF)) with stochastic regression, using haptic information captured from human demonstration. The capabilities of the proposed method are evaluated using two collaborative tasks; co-manipulation of an object (where the human provides the guidance but the robot handles the objects weight) and collaborative assembly of simple interlocking parts. The proposed method is shown to be capable of imitation learning; interpreting human actions and producing equivalent robot motion across a diverse range of initial and final conditions. After verifying that ensemble machine learning can be utilised for real robotics problems, we propose a further extension utilising Weighted Random Forest (WRF) that attaches weights to each tree based on its performance. It is then shown that the WRF approach outperforms RF in HRC tasks.     

面部表情的深度卷积级联森林识别

面部表情识别广泛应用于各种研究领域,针对面部表情识别使用深度神经网络方法结构复杂、可解释性差和传统机器学习方法特征提取缺乏多样性、识别率低的问题.提出了一种新的深度卷积级联森林(Deep Convolution Cascade Forest,DCCF)方法用于人脸面部表情识别,该方法通过卷积神经网络深度学习人脸面部显著特征,并采用基于随机森林的级联结构森林逐层学习识别出不同的面部表情特征,提高了人脸表情的识别准确率.DCCF在JAFFE、CK+和Fer2013 3个公开面部表情数据集进行了实验,并对面部表情提取的5种特征和7种分类方法进行了比较分析,结果显示DCCF在对比的算法中人脸表情识别性能最好,3个数据集的准确率分别达到91.4%,98.7%,71.6%.     

基于排序损失的ECC多标签代码异味检测方法

代码异味是由糟糕的代码或设计问题引起的一种软件特征,严重影响了软件系统的可靠性和可维护性.在软件系统中,一段代码元素可能同时受到多种代码异味的影响,使得软件质量明显下降.多标签分类适用该情况,将高共现的多个代码异味置于同一标签组,可以更好地考虑代码异味的相关性,但现有的多标签代码异味检测方法未考虑同一段代码元素中多种代码异味检测顺序的影响.对此,提出了一种基于排序损失的集成分类器链(ensemble of classifier chains,ECC)多标签代码异味检测方法,该方法选择随机森林作为基础分类器并采取多次迭代ECC的方式,以排序损失最小化为目标,选择一个较优的标签序列集,优化代码异味检测顺序问题,模拟其生成机理,检测一段代码元素是否同时存在长方法长参数列表、复杂类消息链或消息链过大类这3组代码异味.实验采用9个评价指标,结果表明所提出的检测方法优于现有的多标签代码异味检测方法,F1平均值达97.16%.     

基于北斗卫星的森林火灾监控终端设计

针对传统森林火灾监控方式存在的受地形限制,导致火灾发现不及时和定位难的缺陷,设计了基于北斗卫星的森林火灾监控系统,实现了火灾智能识别和预警,并且可以对火源、火场进行快速定位和信息通报;终端包括烟采集处理模块、图像采集模块、DSP处理模块、ARM主控模块和北斗卫星通信模块五部分,从烟和火焰两个方面对森林实时进行监控,分别采用离子烟雾传感器和DSP火焰识别混合算法,提取火焰图像差值和火焰图像边界参数的方法,提高了识别的准确率;试验证明,系统具有结构简单、使用方便和识别智能化的特点。     

A robust critical path in an environment with hybrid uncertainty

In this article, we consider the project critical path problem in an environment with hybrid uncertainty. In this environment, the duration of activities are considered as random fuzzy variables that have probability and fuzzy natures, simultaneously. To obtain a robust critical path with this kind of uncertainty a chance constraints programming model is used. This model is converted to a deterministic model in two stages. In the first stage, the uncertain model is converted to a model with interval parameters by alpha-cut method and distribution function concepts. In the second stage, the interval model is converted to a deterministic model by robust optimization and min-max regret criterion and ultimately a genetic algorithm with a proposed exact algorithm are applied to solve the final model. Finally, some numerical examples are given to show the efficiency of the solution procedure.     

基于随机森林的多站融合项目优先业务筛选机制

在“双循环”新发展格局和“电力数字化”战略中顺势发展“多站融合”项目。基于多站融合布局,提出了一个多站融合项目区域性优先业务筛选的技术方法,从经济发展与能源产业视角,依据各区域属性特征将中国31个省市划分为四类区域,并结合多站融合项目业务清单所梳理的三大类业务范畴,利用随机森林算法的大数据分析法,通过选择区域数据样本,形成了多站融合项目区域性优先业务筛选机制,进而依据优先业务匹配组合预测,实现了各区域优先发展业务的精准甄别和科学选择。实证分析验证了基于随机森林算法的多站融合项目区域性优先业务筛选模型的科学合理性以及有效可靠性,以期为多站融合推进提供有益参考。     

采用RFR-GA算法的薄壁注塑件质量多目标优化

汽车内饰件可由注塑加工获得,但成型过程中塑件产生的翘曲、体积收缩较大,针对该问题,以某汽车薄壁注塑件为例,研究了其注塑工艺参数的优化方法。通过以注塑过程中的最小翘曲和最小体积收缩率为目标函数,以注塑温度、模具温度、注射压力、保压压力、保压时间以及冷却时间等参数作为设计变量,构建了多目标全局优化模型。利用Moldflow软件结合正交试验获得的试验结果训练随机森林回归模型,采用遗传算法对多目标模型进行全局寻优,获得最佳成型工艺参数,即对其成型缺陷进行了优化。结果表明,所提出的优化方法能够得到全局最优解,并同时优化了该汽车薄壁注塑件的翘曲和体积收缩率。将得到的最佳成型工艺参数进行Moldflow试验,可知翘曲和体积收缩率分别优化了74.6%和42.7%。将获得的最佳注塑成型工艺参数进行生产验证,结果表明生产出的薄壁汽车件成型质量较好,满足生产要求。     

Estimation in Random Coefficient Autoregressive Models

Abstract. We propose the quasi‐maximum likelihood method to estimate the parameters of an RCA(1) process, i.e. a random coefficient autoregressive time series of order 1. The strong consistency and the asymptotic normality of the estimators are derived under optimal conditions.     

Thickness and composition dependence of the glass transition temperature in thin random copolymer films

Glass transition temperatures (T_g) of thin poly(styrene-co-methyl methacrylate) and poly(2-vinyl pyridine-co-styrene) films coated on a native oxide surface of Si wafer (100) were measured by ellipsometry. The thickness dependence of T_g can be properly fitted by previously suggested equation developed for homopolymers, based upon a continuous multi-layer model, although one component in thin random copolymer films demonstrates a slightly favorable interaction between a substrate and thin film, and another demonstrates a strongly favorable interaction. Surface and interface have a strong influence on T_g of thin film coated on substrate: the surface has the effect of reducing T_g, whereas the interface increases the T_g according to the degree of interaction between a substrate and thin film. This degree of interaction can be quantified as an interaction parameter (k), and is dependent on the composition of random copolymers. For the estimation of k values of thin random copolymer films, we proposed a parallel type additive function (1/k_ran=w₁/k₁+w₂/k₂) where w is a weight fraction of component.     

森林植被改善对鄱阳湖流域径流和输沙过程的影响

以河流基流为切入点,研究流域植被调节径流、水土保持等微观作用影响大中型河流湖泊径流过程和水沙过程宏观效应的机理。鄱阳湖流域60年来天然降水没有发生趋势性变化。流域森林覆盖率由34.73%上升到63.00%,植被质量改善,赣江等入湖河流基流增加83 m~3/s,河流输沙量减少。2000年以后和2000年以前相比较,枯水期降水径流系数增大,年流量过程平坦化,一定程度上减小洪灾风险,有利于水资源利用和生态环境保护。2001年以后进入鄱阳湖泥沙平均每年减少1 007×10~4t;出湖泥沙增加314×10~4t。因此,鄱阳湖入江水道由淤积转变为冲刷,但出湖流量过程没有趋势性变化。