作者：王武;詹耀清;杨富文;摘要：由于网络传输带宽的限制,在网络传输中可能造成数据的丢失.对于同时具有测量数据和控制数据丢失的一类网络控制系统,研究H_2输出反馈控制问题.数据的丢失采用满足Bernoulli分布的二进制随机变量进行表述.利用矩阵不等式方法给出了H_2动态输出反馈控制器存在的充分条件,所设计的控制器使得闭环系统是均方意义下指数稳定并具有给定的H_2性能.采用SLPMM(Sequentially linear programming matrix method)给出相应的控制器求解算法.最后用数值仿真验证了所提出算法的可行性.

作者：何德峰;季海波;郑涛;摘要：针对未知但有界的持续扰动,提出了一种约束非线性H_∞鲁棒预测控制策略.首先,引入离散系统的输入状态稳定性概念;其次,采用仿射输入定义预测控制的控制律,并给出相应终端约束集的估计解法.进一步,得到预测控制闭环系统的鲁棒稳定性结论.最后,数值仿真验证了上述策略的有效性.

摘要：This paper deals with the optimal robust estima- tion problem for linear uncertain systems with single delayed measurement.The optimal robust estimator is derived based on the reorganized innovation analysis approach.The calculation of the optimal robust estimator involves solving two Riccati dif- ference equations of the same dimensions as that of the original systems and one Lyapunov equation.A numerical example is given to show the effectiveness of the proposed approach.

作者：胡昌华;刘丙杰;摘要：潜在问题是影响大型复杂系统安全性、可靠性的重要因素.神经网络是一种新的潜在问题分析方法,但是其分析结果难以解释.本文提出了一种基于电路结构的神经网络模型(Neural network model based on circuit architecture,CArNN),将CArNN作为个体进行集成,形成神经网络集成用于潜在问题分析.对CArNN模型的鲁棒性进行了分析,提出了两个保证模型鲁棒性的约束条件.利用此方法对一个经典电路进行了分析,结果显示,此方法对潜在电路的正确识别率达到94%.

作者：李航;李敏强;寇纪淞;摘要：多模态函数一般存在多个局部极值解,局部极值解处适应值的大小很大程度上影响了它们被遗传算法搜索到的概率.为了弄清楚这种影响机制,通过分析基因池遗传算法的无限种群动力系统,刻画了双峰函数局部极值解的适值差与系统不动点之间的解析关系,进一步分析推广了理论结果的适用范围.最后,提出针对多模态优化问题的两阶段遗传算法,给出了应用理论结果改善遗传搜索性能的范例,实验结果表明该算法对多模态函数的搜索性能有明显改善,从侧面证明了理论结果在实际应用中的正确性.

摘要：A support vector machine(SVM)approximate-based internal model control(IMC)strategy is presented for the steam valving control of synchronous generators.The proposed SVM IMC strategy includes two main parts:SVM approximate inverse controller and uncertainty compensation in the internal model structure.The SVM inverse controller is derived directly using an input-output approximation approach via Taylor expansion,and it is implemented through nonlinear system identification without further online training.Furthermore,a robustness filter is used for uncertainty compensation in the internal model structure. Simulations show the effectiveness of the SVM IMC strategy for the steam valving control.

摘要：This paper considers the decentralized adaptive regulation via output-feedback for nonlinear systems with integral input- to-state stable(iISS)inverse dynamics,nonlinear uncertainties,and unknown control direction.It is shown that all the signals in the closed-loop system obtained are bounded,and the asymptotic regulation is achieved.A numerical example shows the effectiveness of the design.

摘要：An on-the-fly,self-localization system is developed for mobile robot which is operative in a 3D environment with elaborative 3D landmarks.The robot estimates its pose recursively through a MAP estimator that incorporates the information collected from odometry and unidirectional camera.We build the nonlinear models for these two sensors and maintain that the uncertainty manipulation of robot motion and inaccurate sensor measurements should be embedded and tracked throughout our system.We describe the uncertainty framework in a probabilistic geometry viewpoint and use unscented transform to propagate the uncertainty,which undergoes the given nonlinear functions.Considering the processing power of our robot,image features are extracted in the vicinity of corresponding projected features.In addition,data associations are evaluated by statistical distance. Finally,a series of systematic experiments are conducted to prove the reliable and accurate performance of our system.

作者：周波;韩建达;摘要：用于非线性椭球估计的扩展集员算法在实际应用中存存着数值稳定性差、计算复杂度高以及滤波器参数难以选择等问题.本文提出了一种基于UD分解的自适应扩展集员估计算法,用于解决非线性系统时变状态和参数的联合估计和定界问题.新算法将UD分解与序列更新和选择更新策略结合起米,改进了传统扩展集员算法的数值稳定性和实时性能;同时,对滤波器参数进行自适应选择以进一步降低计算复杂度并达到次优估计结果.仿真实验表明了该算法的有效性和鲁棒性.

摘要：We can usually determine the pose of objects from three lines in a general position.The configuration of three non- coplanar lines that intersect at two points has some particular characteristics,which three lines in a general position do not have. Here,we present a new method of determining object pose using this particular line configuration.In theory,this method enriches the pose estimation methods from three line correspondences.In addition,it provides guidance for practical applications.Furthermore, we propose a method to deal with multi-solution phenomenon and a new iterative method.Simulation results demonstrate that our algorithm works speedily and robustly.