MPC for stable linear systems with model uncertainty

In this paper, we developed a model predictive controller, which is robust to model uncertainty. Systems with stable dynamics are treated. The paper is mainly focused on the output-tracking problem of a system with unknown steady state. The controller is based on a state-space model in which the output is represented as a continuous function of time. Taking advantage of this particular model form, the cost functions is defined in terms of the integral of the output error along an infinite prediction horizon. The model states are assumed perfectly known at each sampling instant (state feedback). The controller is robust for two classes of model uncertainty: the multi-model plant and polytopic input matrix. Simulations examples demonstrate that the approach can be useful for practical application.     

Lower order generalized aggregated model and suboptimal control

A method for reducing the order of a linear time-invariant dynamic system is presented. It is shown that it is possible to retain the predominant eigenvalues (or any other set of eigenvalues) of the exact system in the lower order model that possesses the property that its state is an aggregation of the state variables of the original system. Also it is shown that the output of the reduced order model can be constrained to contain all the modes of the exact output and to be close to the actual output of the original system within a specified tolerance. The performance of the original system is investigated for an optimal output regulator problem, when it is controlled on the assumption that its behavior is governed by that of the lower order model. Relations are obtained for the performance degradation that results with the above suboptimal control policy. Numerical examples show that the suboptimal control can be used in practice to lessen the computational complexity required for the higher order optimal control. The stability of the suboptimal control is not guaranteed; however, it is reasonable to expect it to be asymptotically stable when the order of reduction is not excessively high, because the outputs of the exact and lower order models are tolerably close.     

基于延迟回声状态网的光伏电池板温度预测方法

光伏电池温度变化影响光伏系统输出的稳定性, 精准地预测光伏电池板温度的变化趋势, 对光伏系统智能运行具有重要意义. 为了更好地预测温度的变化趋势, 本文考虑了光伏电池板温度的迟滞效应, 将先前的温度输出作为延迟项引入回声状态网中, 提出了一种基于延迟回声状态网的光伏电池板温度预测模型. 给出一个延迟回声状态网具有回声状态特性的判定条件, 使得预测模型能够稳定地预测光伏电池板温度. 同时, 建立了一套光伏多传感器监测系统, 利用该监测系统采集的数据, 训练和验证模型的准确性. 与回声状态网(Echo state network, ESN), Leaky ESN (Leaky-integrator ESN)和VML ESN (ESN with variable memory length)相比, 仿真结果表明, 本文所提出的延迟回声状态网具有更好的预测性能, 平均绝对百分比误差甚至达到3.45%.     

软件中的错误传播研究

随着计算机技术的广泛应用，计算机软件的安全性显得越来越重要。为了保证软件的安全可靠，必须了解软件可能的错误行为，井采取相应的补救措施来提高其性能。本文采用了基于软件故障注入的方法，对计算机软件中数据状态错误的传播行为进行了研究。结果发现，在给定位置注入的数据状态错，要么所有的都会传播到输出，要麽一个也不会传播到输出。这表明软件测试中的大量数据状态的错误行为，可以通过少量的数据状态错误行为来进行模拟。     

Radial basis function neural network for approximation andestimation of nonlinear stochastic dynamic systems

This paper presents a means to approximate the dynamic and static equations of stochastic nonlinear systems and to estimate state variables based on radial basis function neural network (RBFNN). After a nonparametric approximate model of the system is constructed from a priori experiments or simulations, a suboptimal filter is designed based on the upper bound error in approximating the original unknown plant with nonlinear state and output equations. The procedures for both training and state estimation are described along with discussions on approximation error. Nonlinear systems with linear output equations are considered as a special case of the general formulation. Finally, applications of the proposed RBFNN to the state estimation of highly nonlinear systems are presented to demonstrate the performance and effectiveness of the method.     

Data-based construction of feedback-corrected nonlinear prediction model using feedback neural networks

We propose to fit a recurrent feedback neural network structure to input–output data through prediction error minimization. The recurrent feedback neural network structure takes the form of a nonlinear state estimator, which can compactly represent a multivariable dynamic system with stochastic inputs. The inclusion of the feedback error term as an input to the model allows the user to update the model based on feedback measurements in real-time uses. The model can be useful in a variety of applications including software sensing, process monitoring, and predictive control. A dynamic learning algorithm for training the recurrent neural network has been developed. Through some examples, we evaluate the efficacy of the proposed method and the prediction improvement achieved by the inclusion of the feedback error term.     

Applying a dynamic testability technique to debugging certain classes of software faults

Testability, the tendency for software to reveal its faults during testing, is an important issue for verification and quality assurance. But testability can also be used to good advantage as a debugging technique. Although this concept is more general, we will illustrate it with a specific example: propagation analysis.Propagation Analysis (PA) is a technique for predicting the probability that a data state error affects program output. PA is a technique that produces information about a piece of software's testability. PA bases its prediction on empirical measurement of the probability that an artificial data state error affects program output. After obtaining propagation analysis information for a program and obtaining a failure probability estimate for the program during execution we build a model that can be used to identify possible sites of missing-assignment faults of the form x f(x). Thus we can apply the testability technique PA as a debugging tool.This work supported by a National Research Council NASA-Langley Resident Research Associateship and NASA-Langley Grant NAG-1-884.     

面向数据挖掘的静态电源综合故障诊断研究

静态电源是机场桥载设备最重要的组成部件之一,应用非常广泛。然而,其产生的频发故障会造成设备利用率低、修复率时长和经济损失等问题,在基于桥载设备的安监系统上,提出了静态电源综合故障诊断方法,通过数据挖掘软件,建立了静态电源故障诊断预测模型。通过在线数据库测试结果表明得到了综合故障诊断方法在预测静态电源故障上显现的特点,得到了静态电源的未来状态,实现了对静态电源的实时故障进行预测,进而为解决故障提供方向和目标,最终达到降低经济损失最大化的目的。     

广义联邦滤波器的全局最优性

基于矩阵理论和信息分配原理导出集中卡尔曼滤波、分散化滤波和联邦滤波之间的解析关系, 证明联邦滤波只有当其主滤波器和局部滤波器的维数都相同时, 其全局滤波才是最优的, 并用信号流图直观清晰地说明联邦滤波较分散化滤波结构更简单, 计算量小. 当联邦滤波的主滤波器和局部滤波器的维数不相同时, 只能得到次优解. 文中提出一种广义联邦滤波器的结构, 按信息分配原理重置其一步预测状态误差信息阵和一步预测状态, 获得全局滤波次优解, 并进一步利用全局滤波次优解作为观测量, 反馈修正其一步预测状态得到全局滤波最优解. 文中对最优反馈增益矩阵进行了数学推导, 从理论上证明其滤波结果同集中卡尔曼滤波是等价的, 并通过一个双SINS/GPS组合导航系统的仿真实验结果验证了算法的有效性.     

一种错误率可控的混沌时间序列区间预测算法

针对高风险背景下的混沌时间序列区间预测问题,首次将回声状态网络与一致性预测框架相结合,提出基于两者的混沌时间序列区间预测算法.该算法将回声状态网络的拟合能力与一致性预测区间的可靠性相结合,使得最终的预测区间包含被预测值的频率或概率可以被显著性水平参数所控制,即预测区间具有极高的可信度.同时,由于使用岭回归学习回声状态网络的输出权重,使得算法在学习阶段对样本的留一交叉估计可以被快速地计算,极大地缩短了一致性预测的学习时间.理论分析表明,所提出算法的时间复杂度等价于原始回声状态网络算法的时间复杂度,即算法具有较快的计算速度.实验表明,所提出算法能够较精确地控制预测的错误率,对噪声具有鲁棒性,且预测区间比基于高斯过程的预测区间更加准确地刻画了被预测值的波动范围.     

基于输出误差模型优化的甲板运动预报算法研究

本文提出了一种适用于多种复杂海况的大型舰船甲板运动预报方法,目的在于提高算法对不同海域复杂海况的适用性,以及对甲板运动模型的辨识精度与预报精度。该方法通过将量测数据的时间滞后处理引入输出误差模型来描述甲板运动的动力学模型,引入定阶准则确定了模型最优阶数数对。在此基础上应用了辅助模型递推最小二乘算法进行系统参数辨识并估计输出误差模型中的状态变量。实验结果表明,本文所提出的预报方法在系统参数辨识阶段可以将递推最小二乘算法的辨识精度提高5.13%,并且在预报阶段可以有效地将甲板运动的幅值与相位预测精度提高3.17%。该方法在复杂海况下具备良好的预测性能,适用于大型舰船甲板运动预报。     

Adaptive output feedback control of uncertain nonlinear systems using single-hidden-layer neural networks

We consider adaptive output feedback control of uncertain nonlinear systems, in which both the dynamics and the dimension of the regulated system may be unknown. However, the relative degree of the regulated output is assumed to be known. Given a smooth reference trajectory, the problem is to design a controller that forces the system measurement to track it with bounded errors. The classical approach requires a state observer. Finding a good observer for an uncertain nonlinear system is not an obvious task. We argue that it is sufficient to build an observer for the output tracking error. Ultimate boundedness of the error signals is shown through Lyapunov's direct method. The theoretical results are illustrated in the design of a controller for a fourth-order nonlinear system of relative degree two and a high-bandwidth attitude command system for a model R-50 helicopter.     

On the state observation and output feedback problems for nonlinear uncertain dynamic systems

In this paper, we examine the problems of state observation and state trajectory control by output feedback for the class of non-linear systems in [3,11]. We begin by modifying a known discontinuous variable structure type observer into a continuous type observer that guarantees the observation error is Globally Exponentially Stable (GES). We then modify a known discontinuous variable structure type output feedback controller into continuous type output feedback controller that forces the state to the origin in the GES sense. Specific time-varying bounds on the observation error and the state trajectory are also developed; revealing how the corresponding observer or control parameters can be adjusted to improve the system performance.     

On the optimality of output inventory decisions of a firm under imperfect stochastic markets

A class of suboptimal strategies in price, output and inventories is proposed for a firm in imperfectly competitive stochastic markets. When the demand faced is incompletely known and inventories are held partly as a buffer against fluctuations in demand, and partly as active decision variables, the solution is characterized by maximizing expected profits over two periods with an additive error in the demand function having a known prior density. Simulations of the solution profiles show that for the two-period moving horizon model proposed here, the variance of profits and revenues are sharply reduced, a feature that is most attractive for a risk averse firm.     

Identification of unstable systems using output error andBox-Jenkins model structures

It is well known that the output error and Box-Jenkins model structures cannot be used for prediction error identification of unstable systems. The reason for this is that the predictors in this case generically will be unstable. Typically, this problem is handled by projecting the parameter vector onto the region of stability, which gives erroneous results when the underlying system is unstable. The main contribution of this work is that we derive modified, but asymptotically equivalent, versions of these model structures that can also be applied in the case of unstable systems     

机场巴士运行过程子空间建模与优化

针对机场巴士运行过程影响因素复杂、难以预测运行时间的问题,建立了一种基于子空间辨识算法的机场巴士运行时间预测模型.首先根据运行过程中所产生的多源大数据,考虑不同时段的乘坐人数、发车间隔、道路拥挤度等因素,建立机场巴士运行过程状态空间模型;然后提取适合描述机场巴士运行过程的特征变量作为模型的输入输出,通过子空间辨识方法对模型进行求解;最后以首都机场巴士的一条实际运营路线作为案例进行仿真分析.计算结果表明,该模型预测平均绝对百分误差和均方误差分别为2.25％和4.77,表现均好于传统的BP神经网络预测模型和最小二乘法辨识模型,具有较好的预测精度,有一定的实际应用价值.     

基于可自动扩展的LSTM模型的航空发动机剩余寿命预测方法

对航空发动机进行实时状态监测和健康管理可以有效降低发动机故障风险,确保飞机飞行安全。准确预测航空发动机的剩余寿命是有效监测发动机运行状态的一种重要手段,其中长短期记忆（long-short term memory,LSTM）网络常被使用。但由于航空发动机复杂的机械结构与运行模式,使用传统的LSTM网络对航空发动机的剩余寿命进行单次预测后,所得预测结果的准确率不足以满足其寿命预测的精度要求。基于LSTM网络的广泛使用以及它对时间序列数据的有效预测能力,考虑到采用多级预测的方法能够有效降低预测误差,提出了一种新型的可自动扩展的长短期记忆（automatically expandable LSTM,AELSTM）预测模型。AELSTM模型依托多个子模块逐级连接的网络结构,不断地提取前一级模块的输出误差作为后一级模块的训练值,形成了误差的多级预测机制,有效降低了模型的预测误差,提升了预测结果的准确性。基于美国国家航空航天局发布的C-MAPSS数据集的四个子集对AELSTM模型的预测效果进行了测试,实验结果表明,与传统的LSTM网络相比,AELSTM模型在四个子集上的均方根误差平均减少了95.44%,同时它的预测效果也优于现有的一些先进算法。实验充分验证了AELSTM模型在提升航空发动机剩余寿命预测准确度方面的有效性及优势。     

废气氧传感器Hammerstein模型结构的确定

研究了废气氧(EGO)传感器Hammerstein模型结构辨识方法。静态非线性函数选用双曲正切与多项式组合形式,动态线性环节分别选用带外生变量的自回归(ARX)模型、输出误差(OE)模型和Box-Jenkins(BJ)模型结构。采用交叉准则法进行参数估计和阶次选择,通过仿真比较对模型进行检验。结果表明:最终输出误差(FOE)准则和最终预报误差(FPE)准则均适用于用估计数据选择阶次,但前者比后者更可靠。基于预测误差法的3阶OE模型和BJ模型均可用于EGO传感器Hammerstein模型动态线性环节的建模。     

White noise estimators for networked systems with packet dropouts

This paper studies the optimal and suboptimal deconvolution problems over a network subject to random packet losses, which are modeled by an independent identically distributed Bernoulli process. By the projection formula, an optimal input white noise estimator is first presented with a stochastic Kalman filter. We show that this obtained deconvolution estimator is time-varying, stochastic, and it does not converge to a steady value. Then an alternative suboptimal input white-noise estimator with deterministic gains is developed under a new criterion. The estimator gain and its respective error covariance-matrix information are derived based on a new suboptimal state estimator. It can be shown that the suboptimal input white-noise estimator converges to a steady-state one under appropriate assumptions.