模型参数失配有界下的扩展集员估计方法

在非线性模型参数失配下,直接采用滤波算法很难获到理想的估计状态.本文基于扩展集员估计方法,在状态估计中引入参数的不确定信息,提出一种参数失配有界下的状态估计方法.该方法应用区间或集合运算的法则,计算由参数失配引起的偏差范围,并将其用椭球集外包.在状态估计的预测步,通过该偏差椭球集与先验椭球区间的并运算,得到预测椭球区间;在状态估计的更新步,利用观测椭球集对预测椭球区间进行更新,从而得到后验椭球集合以及状态估计值.最后,在数值仿真和发酵模型中的仿真应用验证了算法的有效性.     

非线性系统中心差分集员估计方法

用于非线性椭球估计的扩展集员算法在实际应用中存在着实现性差、边界估计相对保守等缺陷.本文提出了一种用于非线性系统状态估计的中心差分集员估计方法,以改善传统非线性集员滤波算法的估计性能.为克服泰勒展开的固有缺陷,采用低阶多维Stirling内插多项式代替泰勒展开实现非线性模型的线性化处理;利用半定规划方法对线性化误差进行外包定界并将其融入过程噪声和量测噪声中,以降低误差定界的保守性;量测更新中,为提高算法的实时性,将量测椭球松弛为多个带的交,依次参与状态椭球的更新,从而实现状态定界椭球的次优估计;同时,对椭球—带交集迭代过程中椭球中心到超平面的归一化距离的计算方法进行了改进,使当前时刻每次迭代的椭球均参与计算并选取最优值,以减小累计误差.仿真结果表明了本文所提出算法的有效性和改进性能.     

一种新的基于保证定界椭球算法的非线性集员滤波器

基于未知但有界噪声假设的集员滤波器为传统的概率化滤波方法提供了一种可行的替代选择, 然而其潜在的计算负担和保守性考虑制约了该方法的实际应用. 本文提出一种新的基于保证定界椭球近似的改进集员滤波方法, 用于解决针对非线性系统的状态估计问题,在保证实时性的前提下降低了算法的保守性. 首先,对非线性模型进行线性化处理,采用DC (Difference of convex)规划方法对线性化误差进行外包定界, 并通过椭球近似将其融合到系统噪声中; 在此基础上提出了一种结合了椭球直和计算和基于迭代外定界椭球算法的椭球--带交集计算 所构成的经典预测--更新步骤来估计得到状态的可行椭球集. 与常规的非线性扩展集员滤波方法的仿真比较表明了本文所提出算法的有效性和改进性能.     

基于MIT规则的自适应扩展集员估计方法

用于非线性椭球估计的自适应扩展集员(Adaptive extended set-membership filter, AESMF)算法在实际应用中存在着过程噪声设定椭球与真实噪声椭球失配的问题, 导致滤波器的估计出现偏差甚至发散. 本文提出了一种基于MIT规则过程噪声椭球最优化的自适应扩展集员估计算法(MIT-AESMF), 用于解决非线性系统时变状态和参数的联合估计和定界中过程噪声无法精确建模问题的新算法. 本算法通过MIT优化规则,在线计算使一步预测偏差包络椭球最小化的过程噪声包络椭球, 以此保证滤波器健康指标满足有效条件; 最后, 采用地面移动机器人状态和动力学参数联合估计验证了所提出方法的有效性.     

整车主动悬架系统分布式滚动时域一致性估计

考虑整车主动悬架系统的约束状态估计问题,本文提出基于一致性原理的分布式滚动时域估计(DMHE)算法.首先,为了降低状态估计过程中的计算量,将整车主动悬架系统分解为若干降阶子系统.其次,为提高分布式状态估计效果,采用滚动时域估计(MHE)方法处理主动悬架系统的状态和噪声约束.考虑子系统与邻居估计状态的相关性,在采样间隔中执行多次一致性原理实现主动悬架系统状态的信息融合,进一步建立了算法的稳定性充分条件.最后,通过对比仿真实验验证算法的有效性和优越性.     

传感器网络中鲁棒状态信息融合抗差卡尔曼滤波器

研究了无线传感器网络中的分布式鲁棒状态信息融合问题. 在局部状态估计层, 基于鲁棒统计学理论提出了适用于噪声相关情况的抗差(扩展)卡尔曼滤波器. 在融合中心层, 针对局部估计相关未知性和不完整性, 给出了不依赖于互协方差阵的稳健航迹融合方法—–内椭球逼近法. 仿真结果证实了算法的有效性: 所提出的抗差卡尔曼滤波器在野值存在情况下, 性能退化远低于传统卡尔曼滤波器(28.6%比428.6%); 所提出的内椭球逼近法获得比协方并交叉法更好的融合估计性能, 且不需要局部估计相关性的先验知识.     

通讯约束下的线性系统状态降维与估计

本文面向状态估计, 考察了通讯功率受限时线性动态系统状态的降维问题. 为了满足平行信道传输数据的维数限制和通讯功率约束, 采取降低状态维数的方法, 通过传输信号的新息, 提高传输效率, 利用有限的通信资源, 使得接收端的状态估计达到最优. 本文采用差分脉冲编码调制系统(DPCM), 基于最小误差熵估计准则和Kalman估计算法, 得出了最优的状态降维矩阵的设计方法, 并且对随机系统的可估计性以及对相应确定性系统的能观性进行了分析. 分析和仿真结果表明, 这种设计方法在传输信号满足通讯功率限制的条件下可以使接收端的状态估计性能达到最优.     

基于改进集员滤波的港口自动跨运车状态估计方法

针对自动跨运车状态估计问题,设计改进的集员滤波算法,在未知有界噪声环境下,获取自动跨运车实时运动状态的估计信息.首先,将自动跨运车运动学模型进行线性化处理,同时考虑其转向因素和侧倾因素,得到车辆的动力学线性模型;其次,将可能存在的内外部扰动建模为未知有界噪声,进而设计改进的集员滤波器,通过获取状态椭球域实现对自动跨运车运动参数的状态估计,同时给出改进的集员滤波算法;最后,通过仿真实验验证所提出算法的可行性和有效性.实验结果表明,所提出的改进集员滤波算法具有良好的状态估计性能.     

量子系统中状态估计方法的综述

从广泛用于实验量子领域的典型状态估计方法,到基于系统论观点、可用于量子反馈控制的状态估计方法,详细综述了4种测量方式下的相应量子状态估计方法及其适用背景.通过其发展历程的叙述,从本质上阐述了估计的基本原理,从技术上对各种方法进行了相应的分析和比较.同时,对量子状态估计和经典状态估计进行了相应的比较,并对量子系统中的状态估计方法作了总结.     

改进强跟踪滤波算法及其在汽车状态估计中的应用

准确实时地获取汽车行驶过程中的状态变量,对汽车底盘控制有着重要的意义,而这些关键状态往往难以直接测量或 者成本较高.结合纵向、侧向和横摆三自由度非线性汽车模型,将改进强跟踪滤波(Improved strong track filter, ISTF)算法应用到汽车的状态估计中,并改进了算 法的稳定性.与扩展卡尔曼滤波(Extended Kalman filter, EKF)算法进行比较分析.通过Carsim和Matlab/Simulink联合仿真和实车双移线实验验证算法,结果 表明,该算法在估计精度、跟踪速度、抑制噪声等方面均优于扩展卡尔曼滤波算法,满足汽车状态估计器的软件性能要求.     

State estimation in the presence of bounded disturbances

This contribution proposes a robust recursive algorithm for the state estimation of linear models with unknown but bounded disturbances corrupting both the state and measurement vectors. A novel approach based on state bounding techniques is presented. The proposed algorithm can be decomposed into two steps: time updating and observation updating that uses a switching estimation Kalman-like gain matrix. Particular emphasis will be given to the design of a weighting factor that ensures the stability of the estimation error.     

改进的全对称多胞形集员状态估计算法

针对带有有界的噪声和参数的非线性离散时间系统, 提出了一种改进的全对称多胞形集员状态估计算法.在算法的时间更新过程中, 采用区间算术的方法计算一个包含系统轨迹的全对称多胞形.在算法的量测更新过程中, 则要首先在状态空间中给出一个集合作为与量测输出相一致的区域的外界描述, 然后计算一个具有最小容积的全对称多胞形作为时间更新全对称多胞形与此集合的交集的外界描述.由于此集合可表示为多个带的交集, 所以需要研究全对称多胞形与带的交集的外界描述方法.在提出改进的外界描述方法之前, 指出了原始外界描述方法的保守性.改进的外界描述方法给出了新的包含二者交集的全对称多胞形族, 然后找到具有最小容积的全对称多胞形作为二者交集的外界描述. 此后证明了改进外界描述方法得到的全对称多胞形不会比原始方法大. 最后, 采用仿真实验来检验不同噪声分布对算法性能的影响.仿真结果表明了改进算法得到的状态估计的均方误差和全对称多胞形的容积比原始算法小, 而且当存在重尾分布噪声时此优势更加明显.     

Adaptive variable structure state estimation for uncertain systems with persistently bounded disturbances

This paper develops an adaptive state estimator design methodology for nonlinear systems with unknown nonlinearities and persistently bounded disturbances. In the proposed estimation scheme, the boundary layer strategy in variable structure techniques is utilized to design a continuous state estimator such that the undesirable chattering phenomenon is avoided; and the adaptive bounding technique is used for online estimation of the unknown bounding parameter. The existence condition of the adaptive estimators is provided in terms of linear matrix inequality (LMI). Since the orthogonal projection of the state estimation error onto the null space of the linear measurement distribution matrix is used in the derivation process, the update law of bounding parameter estimate is represented in terms of the available measurement error. The proposed estimator can ensure that the state estimation error is uniformly ultimately bounded (UUB) with an ultimate bound. Furthermore, using the existing LMI optimization technique, a suboptimal adaptive state estimator can be obtained in the sense of minimizing an upper bound of the peak gains in the ultimate bound. Finally, a simulation example is given to illustrate the effectiveness of the proposed design method. Copyright © 2010 John Wiley & Sons, Ltd.     

Improving continuous–discrete interval observers with application to microalgae-based bioprocesses

Based on intervals bounding the uncertain initial conditions, inputs, model parameters and measurements, interval observers provide guaranteed intervals for the state trajectory. However, most of the published studies focus on methods relying on continuous-time on-line measurements, or at least, relatively fast sampling. In this study, interval state estimation methods are proposed in the situation, quite common in biological systems, where measurements are only available at discrete, and possibly rare, times. The attention is focused on defining predictors preserving the boundedness of the state variables between two measurement times assuming bounded uncertainties. The methods are tested with simulated data from continuous cultures of green algae.     

Reachable set estimation for switched positive systems

This paper focuses on the problem of reachable set estimation for discrete-time switched positive systems under two possible classes of exogenous disturbance. The multiple linear copositive Lyapunov function approach is applied to determine the bounding hyper-pyramids for the reachable set. Based on some Lyapunov-based inequalities and the linear version of the S-procedure technique, the bounding hyper-pyramids for the reachable set can be determined by solving a set of inequalities. Two optimisation methods are adopted to make the bounding hyper-pyramids as small as possible. Genetic Algorithm (GA) is utilised to search for the optimal value of the decision variables in the obtained inequalities. Finally, numerical examples are given to illustrate the effectiveness of the theoretical findings.     

YOLOv3与顶点偏移估计相结合的车牌定位

深层卷积神经网络(deep convolutional neural networks, DCNN)因其能够自动学习图像有效特征,被广泛应用于视觉目标检测.为克服DCNN目标检测算法大多因采用矩形检测框,而无法有效地应对非约束环境下倾斜性车牌的准确定位问题.提出一种可同时输出矩形目标检测框与关键点的车牌定位解决方案,并具体以YOLOv3所用网络为对象,通过扩展其输出维度,增设车牌顶点相对于矩形检测输出框角点的偏移量损失,在保留其高效计算性能的前提下,训练使其可同时输出矩形检测框及车牌顶点,实现精准定位.在广泛使用的大型非约束性车牌数据集CCPD上的实验结果显示,所提算法不仅可以准确检测车牌顶点,而且能够在Base,Tilt和Weather子集上取得99%以上的定位精度.该方法还可扩展至其他需同时输出目标检测框及关键点的应用领域,具有较好的应用价值.     

A Real-Time Multi-Stage Architecture for Pose Estimation of Zebrafish Head with Convolutional Neural Networks

In order to conduct optical neurophysiology experiments on a freely swimming zebrafish,it is essential to quantify the zebrafish head to determine exact lighting positions.To efficiently quantify a zebrafish head's behaviors with limited resources,we propose a real-time multi-stage architecture based on convolutional neural networks for pose estimation of the zebrafish head on CPUs.Each stage is implemented with a small neural network.Specifically,a light-weight object detector named Micro-YOLO is used to detect a coarse region of the zebrafish head in the first stage.In the second stage,a tiny bounding box refinement network is devised to produce a high-quality bounding box around the zebrafish head.Finally,a small pose estimation network named tiny-hourglass is designed to detect keypoints in the zebrafish head.The experimental results show that using Micro-YOLO combined with RegressNet to predict the zebrafish head region is not only more accurate but also much faster than Faster R-CNN which is the representative of two-stage detectors.Compared with DeepLabCut,a state-of-the-art method to estimate poses for user-defined body parts,our multi-stage architecture can achieve a higher accuracy,and runs 19x faster than it on CPUs.