基于模型预测—中枢模式发生器的六足机器人轨迹跟踪控制

为了模仿动物卓越的运动能力和环境适应能力,提出了六足仿生机器人的轨迹跟踪控制方法。首先建立了机器人的运动学模型,接着通过转向参数将机器人的速度和角速度与中枢模式发生器（CPG）参数结合起来,设计了转换函数。然后通过转换函数将模型预测控制器和CPG网络结合起来,提出了基于CPG的模型预测控制器（MPC-CPG）,并证明了其稳定性。最后对机器人跟踪圆周轨迹和直线轨迹进行了仿真和实验。实验表明,在有初始误差的条件下,机器人在MPC-CPG控制器的作用下能够快速地消除位置误差和航向角误差,跟踪上参考轨迹。轨迹跟踪的位置误差始终保持在-0.1～0.1 m,航向角误差保持在-27?～20?。在MPC-CPG控制器的作用下,机器人不仅具有较高的轨迹跟踪精度,同时还表现出良好的运动平滑性和协调性,进一步验证了所提出的MPC-CPG控制器的有效性。     

Inverse-compensation-based DETC for nonlinear hysteresis system with disturbance

To control a nonlinear system with both hysteresis and disturbance, it is necessary to establish a hysteresis model and improve the disturbance rejection ability. However, the input signal implicitly involved in the classical hysteresis model can lead to difficulty in constructing a compensator. In this study, a hysteresis model in explicit form is proposed with a bounded auxiliary variable. Then, a model-based inverse is constructed for approximate compensation for the hysteresis. Moreover, the compensation error, which is considered a part of the disturbance, is proved to be bounded. Disturbance estimation triggered control (DETC) is utilized to address the compensation error coupled with the external disturbance. According to the disturbance effect indicator (DEI), DETC can improve the system control performance by considering the disturbance effect judgment. Experimental results are presented to illustrate the potential of the proposed technique.     

特征挖掘与区域增强的弱监督时序动作定位

弱监督时序动作定位旨在定位视频中行为实例的起止边界及识别相应的行为。现有方法尽管取得很大进展,但依然存在动作定位不完整及短动作的漏检问题。为此,提出了特征挖掘与区域增强（FMRE）的定位方法。首先,通过基础分支计算视频片段之间的相似分数,并以此分数聚合上下文信息,得到更具有区别性的段分类分数,实现动作的完整定位;然后,添加增强分支,对基础分支定位中持续时间较短的动作提案沿时间维度进行动态上采样,进而采用多头自注意机制对动作提案间的时间结构显式建模,促进具有时间依赖关系的动作定位且防止短动作的漏检;最后,在两个分支之间构建伪标签互监督,逐步改进在训练过程中生成动作提案的质量。该算法在THUMOS14和ActivityNet1.3数据集上分别取得了70.3%和40.7%的检测性能,证明了所提算法的有效性。     

少样本深度学习算法赋能二维衍射超材料逆向设计

二维衍射超材料是研究高阶衍射光束的光学手性响应的理想材料。随着计算机软硬件技术的发展,越来越多的深度学习算法应用于超材料这一领域。当前,深度学习算法大多需要大量训练数据作为支撑,造成了较大时间开销,并且其模型的可解释性通常较弱。针对以上情况,本文提出一种基于数据增强的迭代少样本深度学习算法,利用少量的仿真数据,训练正向预测的网络模型,进而生成大量“伪数据”,再使用扩充后的混合数据集进行逆向设计模型训练,通过多次迭代得到超材料结构参数的准确值。与以往大多数深度学习方法相比,本文方法将仿真数据集规模缩小了至少一到两个数量级,有效降低了对仿真光谱的依赖,且设计成功率不低于96%。同时,从数据的角度,根据二维衍射超材料的几何参数对光学响应的影响解释了实验结果,增强了模型的可解释性。结果表明,本文方法为超灵敏、可调控偏振光电子器件的研发提供了新思路。     

一种深度回声状态网络的输入尺度自适应算法

深度回声状态网络是回声状态网络与深度学习思想的结合,合理选取不同谱半径的内部状态矩阵和弱积分参数能有效增强深度回声状态网络的多尺度时域特性。利用数据可视化分析输出矩阵在不同网络层中的分布关系,发现高层网络中部分神经元处于饱和工作状态且该状态抑制了网络动态预测能力。提出一种深度回声状态网络的输入矩阵自适应算法,在对网络内部状态的均值和方差进行递推估计的基础上判断神经元饱和状态,通过自适应调整各层输入权重的值来增强神经元动态性。数值计算结果表明,基于输入尺度自适应算法的深度回声状态网络相对同等规模的单层回声状态网络对于动态系统的预测精度有成倍提升。     

The method of fundamental solutions for problems in static thermo-elasticity with incomplete boundary data

An inverse problem in static thermo-elasticity is investigated. The aim is to reconstruct the unspecified boundary data, as well as the temperature and displacement inside a body from over-specified boundary data measured on an accessible portion of its boundary. The problem is linear but ill-posed. The uniqueness of the solution is established but the continuous dependence on the input data is violated. In order to reconstruct a stable and accurate solution, the method of fundamental solutions is combined with Tikhonov regularization where the regularization parameter is selected based on the L-curve criterion. Numerical results are presented in both two and three dimensions showing the feasibility and ease of implementation of the proposed technique.     

Real‐time updating of structural mechanics models using Kalman filtering,modified constitutive relation error,and proper generalized decomposition

The paper aims at proposing a new strategy for real‐time identification or updating of structural mechanics models defined as dynamical systems. The main idea is to introduce the modified constitutive relation error concept, which is a practical tool that enables to efficiently solve identification problems with highly corrupted data, into the Kalman filtering, which is a classical framework for data assimilation. Furthermore, a PGD‐based model reduction method is performed in order to optimize capabilities of the online updating strategy. Performances of the proposed approach, in terms of robustness gain and computational cost reduction, are illustrated on several unsteady thermal applications. Copyright © 2015 John Wiley & Sons, Ltd.     

Comparison of the identification performance of conventional FEM updating and integrated DIC

Full‐field identification methods are increasingly used to adequately identify constitutive parameters to describe the mechanical behavior of materials. This paper investigates the more recently introduced one‐step method of integrated digital image correlation (IDIC) with respect to the most commonly used two‐step method of finite element model updating (FEMU), which uses a subset‐based DIC algorithm. To make the comparison as objective as possible, both methods are implemented in the most equivalent manner and use the same FE model. Various virtual test cases are studied to assess the performance of both methods when subjected to different error sources: (1) systematic errors, (2) poor initial guesses for the constitutive parameters, (3) image noise, (4) constitutive model errors, and (5) experimental errors. Results show that, despite the mathematical similarity of both methods, IDIC produces less erroneous and more reliable results than FEMU, particularly for more challenging test cases exhibiting small displacements, complex kinematics, misalignment of the specimen, and image noise. Copyright © 2015 John Wiley & Sons, Ltd.     

获取界面响应不可测数据的频响探针关键技术研究

目的针对传统逆子结构理论在求解过程中界面响应难以实测的问题,提出一种利用频响探针技术来获取该界面响应的关键技术。方法首先基于该频响探针的动力学微分方程,从理论上推导了该频响探针技术的理论公式,然后对建立的二级单点刚性耦合系统进行了有限元数值验证,将利用频响探针技术预测得到的难测原点频响函数与有限元计算值进行比较,并将该预测值代入逆子结构理论公式中,得到了部件频响函数的预测值,将该预测值与有限元计算值进行了对比验证。结果预测值与有限元计算值高度吻合,验证了该理论的准确性。结论该频响探针关键技术在获取界面响应不可测数据方面,有很好的应用价值。