Optimal nonuniform sampling interval and test-input design for identification of physiological systems from very limited data

Optimal design of test-inputs and sampling intervals in experiments for linear system identification is treated as a nonlinear integer optimization problem. The criterion is a function of the Fisher information matrix, the inverse of which gives a lower bound for the covariance matrix of the parameter estimates. Emphasis is placed on optimum design of nonuniform data sampling intervals when experimental constraints allow only a limited number of discrete-time measurements of the output. A solution algorithm based on a steepest descent strategy is developed and applied to the design of a biologic experiment for estimating the parameters of a model of the dynamics of thyroid hormone metabolism. The effects on parameter accuracy of different model representations are demonstrated numerically, a canonical representation yielding far poorer accuracies than the original process model for nonoptimal sampling schedules, but comparable accuracies when these schedules are optimized. Several objective functions for optimization are compared. The overall results indicate that sampling schedule optimization is a very fruitful approach to maximizing expected parameter estimation accuracies when the sample size is small.     

动态驱动神经网络辨识永磁直线同步电动机模型

永磁直线同步电动机(PMLSM)模型的建立对研究其稳态特性、动态特性和控制策略都是非常重要的．本文利用动态驱动神经网络对其进行建模,并在代价函数一致的基础上加入残差分析法来辨识模型的阶次,使得神经网络具有自动识别阶次的能力．为了克服神经网络结构依靠人工试凑的不足,使用基于Hession矩阵的修剪法来优化其结构．考虑到改进BP算法(学习速率自适应、动量项的方法)的一些固有缺点,使用NDEKF(基于节点的解耦扩展Kalman滤波器算法)来训练网络．实验证明,混合网络能够准确辨识出试验样机的阶次并且输出结果与实际结果十分接近;同时将NDEKF与改进BP算法进行对比,NDEKF算法具有收敛较快、泛化能力强等特点．     

粒子群优化RBF神经网络光伏电池建模研究

本文研究神经网络在光伏电池建模优化问题。由于光伏电池具有高度非线性特性,其输出功率受到外界自然因素的影响,使得传统方法不能满足光伏控制系统动态要求。针对上述问题,本文提出一种粒子群优化的神经网络光伏电池建模算法。改进的方法以日照、温度和负载电压作为提出的RBF神经网络模型的输入值,把光伏电池的输出功率作为神经网络的输出,采用RBF神经网络对光伏电池进行建模,同时利用粒子群算法对神经网络参数进行优化,最后建立光伏电池的动态响应模型。仿真实验结果证明,所提模型更好地克服传统方法的缺点,收敛速度快,具有较高的预测精度和适合能力。     

神经网络非线性多步预测逆控制方法研究*

提出了基于多步预测控制方法的多变量非线性神经网络逆控制方案。利用预测模型对系统动态特性进行预测,使用一个带有时延因子的前馈神经网络作为控制器,利用多步预测性能指标对其在线训练,实现神经网络逆系统;在多步预测过程中还对每一步的预测误差进行预测,以实现预测误差补偿。将所提出的控制算法用于锅炉这种大滞后非线性对象的控制,仿真实验证明,该控制策略具有良好的解耦和动态跟踪性能。     

Comparative nonlinear modeling of renal autoregulation in rats:Volterra approach versus artificial neural networks

In this paper, feedforward neural networks with two types of activation functions (sigmoidal and polynomial) are utilized for modeling the nonlinear dynamic relation between renal blood pressure and flow data, and their performance is compared to Volterra models obtained by use of the leading kernel estimation method based on Laguerre expansions. The results for the two types of artificial neural networks and the Volterra models are comparable in terms of normalized mean square error (NMSE) of the respective output prediction for independent testing data. However, the Volterra models obtained via the Laguerre expansion technique achieve this prediction NMSE with approximately half the number of free parameters relative to either neural-network model. However, both approaches are deemed effective in modeling nonlinear dynamic systems and their cooperative use is recommended in general     

Piecewise Fuzzy Anti-Windup Dynamic Output Feedback Control of Nonlinear Processes With Amplitude and Rate Actuator Saturations

In this paper, a novel anti-windup dynamic output compensator is developed to deal with the robust H _infin output feedback control problem of nonlinear processes with amplitude and rate actuator saturations and external disturbances. Via fuzzy modeling of nonlinear systems, the proposed piecewise fuzzy anti-windup dynamic output feedback controller is designed based on piecewise quadratic Lyapunov functions. It is shown that with sector conditions, robust output feedback stabilization of an input-constrained nonlinear process can be formulated as a convex optimization problem subject to linear matrix inequalities. Simulation study on a strongly nonlinear continuously stirred tank reactor (CSTR) benchmark plant is given to show the performance of the proposed anti-windup dynamic compensator.     

A new approach to radial basis function-based polynomial neural networks: analysis and design

In this study, we introduce a new topology of radial basis function-based polynomial neural networks (RPNNs) that is based on a genetically optimized multi-layer perceptron with radial polynomial neurons (RPNs). This paper offers a comprehensive design methodology involving various mechanisms of optimization, especially fuzzy C-means (FCM) clustering and particle swarm optimization (PSO). In contrast to the typical architectures encountered in polynomial neural networks (PNNs), our main objective is to develop a topology and establish a comprehensive design strategy of RPNNs: (a) The architecture of the proposed network consists of radial polynomial neurons (RPN). These neurons are fully reflective of the structure encountered in numeric data, which are granulated with the aid of FCM clustering. RPN dwells on the concepts of a collection of radial basis function and the function-based nonlinear polynomial processing. (b) The PSO-based design procedure being applied to each layer of the RPNN leads to the selection of preferred nodes of the network whose local parameters (such as the number of input variables, a collection of the specific subset of input variables, the order of the polynomial, the number of clusters of FCM clustering, and a fuzzification coefficient of the FCM method) are properly adjusted. The performance of the RPNN is quantified through a series of experiments where we use several modeling benchmarks, namely a synthetic three-dimensional data and learning machine data (computer hardware data, abalone data, MPG data, and Boston housing data) already used in neuro-fuzzy modeling. A comparative analysis shows that the proposed RPNN exhibits higher accuracy in comparison with some previous models available in the literature.     

Robust output containment control of multi-agent systems with unknown heterogeneous nonlinear uncertainties in directed networks

Output containment problem for high-order nonlinear time-invariant multi-agent systems in directed networks is investigated in this paper. The output is related with the observation matrix. The dimensions of observation matrix are extended so that it is non-singular. Then the containment problem is transformed into stability problem. The model of each agent is constructed by a nominal system combined with uncertainties. A robust controller, which includes a nominal controller and a robust compensator, is proposed to achieve output containment and restrain external uncertainties. The nominal controller is based on the output feedback and the nominal system constructed by the nominal controller contains desired containment properties. The robust compensator design is based on robust signal compensation technology for restraining the effects of external disturbances. A sufficient condition on the output containment is proposed and the containment errors can be made as small as desired with the expected convergence rate. Finally, numerical simulation is presented to demonstrate the effectiveness of the control method.     

Structure selection of neurofuzzy networks based on support vector regression

Neurofuzzy networks are being used increasingly to model non-linear dynamic systems, since they have the approximating ability of neural networks and the transparency of fuzzy systems. However, good generalization results can only be obtained if the structure of the network is suitably chosen. It is shown here that the structure of neurofuzzy networks with scatter partitioning can be obtained from the support vectors (SV) of the Support Vector Regression (SVR), since the SVR can be transformed to a neurofuzzy network. The main advantage of this approach is that the structure of the neurofuzzy networks can now be objectively chosen, as the SV are obtained by constrained optimization for a given modelling error bound. Since neurofuzzy networks are linear-inweights networks, the estimate of the weights of the networks can be obtained by the linear least-squares method. The properties of neurofuzzy networks based on the SV are derived, and its performance is illustrated by a simulation example involving a nonlinear system, and the modeling of Southern Oscillation Index.     

Nonlinear identification of a gasoline HCCI engine using neural networks coupled with principal component analysis

Homogeneous charge compression ignition (HCCI) is a futuristic combustion technology that operates with high efficiency and reduced emissions. HCCI combustion is characterized by complex nonlinear dynamics which necessitates the use of a predictive model in controller design. Developing a physics based model for HCCI involves significant development times and associated costs arising from developing simulation models and calibration. In this paper, a neural networks (NN) based methodology is reported where black box type models are developed to predict HCCI combustion behavior during transient operation. The NN based approach can be considered a low cost and quick alternative to the traditional physics based modeling. A multi-input single-output model was developed each for indicated net mean effective pressure, combustion phasing, maximum in-cylinder pressure rise rate and equivalent air–fuel ratio. The two popular architectures namely multi-layer perceptron (MLP) and radial basis network (RBN) models were compared with respect to design, prediction performance and overall applicability to the transient HCCI modeling problem. A principal component analysis (PCA) is done as a pre-processing step to reduce input dimension thereby reducing memory requirements of the models. Also, PCA reduces the cross-validation time required to identify optimal model hyper-parameters. On comparing the model predictions with the experimental data, it was shown that neural networks can be a powerful approach for non-linear identification of a complex combustion system like the HCCI engine.     

基于随机配置网络的非线性系统智能建模方法

针对一类非线性离散动态系统,研究非线性系统的智能建模方法.首先,采用带遗忘因子的递推最小二乘法对低阶模型的未知参数进行辨识;然后,对高阶非线性部分采用随机配置网络进行估计;最后,利用两种辨识方法在外部误差准则下对系统进行交替辨识,进而提出一种改进的非线性系统交替辨识的智能建模方法.将随机配置网络与递推最小二乘算法相结合,可有效提高非线性系统的辨识精度,并且通过数值仿真实验进行对比分析以验证所提出算法的有效性.     

RBF-ARX model-based MPC strategies with application to a water tank system

A hybrid pseudo-linear RBF-ARX model that combines Gaussian radial basis function (RBF) networks and linear ARX model structure is utilized for representing the dynamic behavior of a class of smooth nonlinear and non-stationary systems. This model is locally linear at each working point and globally nonlinear within whole working range. Based on the structural characteristics of the RBF-ARX model, three receding horizon predictive control (RBF-ARX-MPC) strategies are designed: (1) the RBF-ARX-MPC algorithm based on single-point linearization (MPC-SPL); (2) the RBF-ARX-MPC algorithm based on multi-point linearization (MPC-MPL); and (3) the RBF-ARX-MPC algorithm based on globally nonlinear optimization (MPC-GNO). In the MPC-SPL, the future multi-step-ahead predictive output of the system is obtained based on the local linearization of the RBF-ARX model at only current working-point, while in the MPC-MPL the future long-term output prediction is obtained according to the future local characteristics from previous online optimization results of the RBF-ARX model based MPC. In the MPC-GNO, the globally nonlinear characteristics of the RBF-ARX model are fully used for online getting control variables of the MPC. Real-time control experiments for the three type MPCs are carried out on a water tank system, which are also compared with a classical PID control and a traditional linear ARX model-based MPC. The results verify that the modeling method and the model-based predictive control strategies are realizable and effective for the nonlinear and unstable system. Moreover, it is also shown that the MPC-GNO can obtain better control performance but need more computation time compared to the other MPCs, which makes it possible to be applied into some slowly varying processes.     

Dynamic neural networks with hybrid structures for nonlinear system identification

Dynamic neural networks (DNNs) have important properties that make them convenient to be used together with nonlinear control approaches based on state space models and differential geometry, such as feedback linearisation. However the mapping capability of DNNs are quite limited due to their fixed structure, that is, the number of layers and the number of hidden units. An example shown in this paper has demonstrated this limitation of DNNs. The development of novel DNN structures, which has good mapping capability, is a relevant challenge being addressed in this paper. Although the structure is changed minorly only, the mapping capability of the new designed DNN in this paper has been improved dramatically. Previous work [J. Deng et al., 2005. The dynamic neural network of a hybrid structure for nonlinear system identification. In: 16th IFAC World Congress, Prague.] presents a new dynamic neural network structure which is suitable for the identification of highly nonlinear systems, which needs the outputs from the real system for training and operation. This paper presents a hybrid dynamic neural network structure which presents a similar idea of serial–parallel hybrid structure, but it uses an output from another neural network for training and operation classified as a serial–parallel model. This type of DNNs does not require the output of the plant to be used as an input to the model. This neural network has the advantages of good mapping capabilities and flexibilities in training complicated systems, compared to the existed DNNs. A theoretical proof showing how this hybrid dynamic neural network can approximate finite trajectories of general nonlinear dynamic systems is given. To illustrate the capabilities of the new structure, neural networks are trained to identify a real nonlinear 3D crane system.     

Object-oriented modeling and simulation of neural networks

The field of neural networks is being investigated by many researchers in order to provide solutions to difficult problems in the area of manufacturing systems. Computer simulation of neural networks is an important part of this investigation. This paper applies concepts from an important trend in software engineering research, namely object-oriented programming, to model neural networks.The design and implementation of a software object library is crucial to obtaining the full benefits of object-oriented programming. In this paper we discuss the design and implementation of a foundation library of software objects for the purpose of simulating and validating different network architectures and learning rules. The library contains objects that implement various types of nodes and learning rules. We discuss the results of our experiments to illustrate the benefits of using an object-oriented approach to modeling neural networks.     

Nonlinear modeling of MCFC stack based on RBF neural networks identification

Modeling molten carbonate fuel cells (MCFC) is very difficult and the most existing models are based on conversation laws which are too complicated to be used to design a control system. This paper presents an application of radial basis functions (RBF) neural networks identification to develop a nonlinear temperature model of MCFC stack. The temperature characters of MCFC stack are briefly analyzed. A summary of RBF neural networks modeling of MCFC is introduced. The simulation tests reveal that it is feasible to establish the model of MCFC stack using RBF neural networks identification. The modeling process avoids using complicated differential equations to describe the stack and the neural networks model developed can be used to predict the temperature responses online which makes it possible to design online controller of MCFC stack.     

Management,control, and design of integrated networks with real-time dynamic routing

Integrated real-time dynamic routing (IRR) networks provide dynamic routing features for multiple classes-of-service on an integrated transport network. In this paper it is shown that IRR networks allow reduced network management costs since with real-time dynamic routing a number of network operations are simplified or eliminated. These simplifications include eliminating the storage of voluminous routing tables in the network switches, eliminating the calculation of routing tables in network design, simplifying the routing administration operations which require downloading new routing information to the network, and eliminating the automatic rerouting function in on-line traffic management. A new bandwidth allocation technique is described here which is based on the optimal solution of a network bandwidth allocation model for IRR networks. The model achieves significant improvement in both the average network blocking and node pair blocking distribution when the network is in a congested state such as under peak-day loads. In a paper to appear in the next Journal issue we further describe a new algorithm for the transport design of IRR networks which achieves near-optimal capacity engineering. These optimization techniques attain significant capital cost reductions and network performance improvements by properly modeling the more efficient operation of IRR networks.     

Transport network design of integrated networks with real-time dynamic routing

Integrated real-time dynamic routing (IRR) networks provide dynamic routing features for multiple classes-of-service on an integrated transport network. In a previous Journal paper it is shown that IRR networks allow reduced network management costs since with real-time dynamic routing a number of network operations are simplified or eliminated, leading to savings in operations costs and expenses. In this paper a new algorithm is described for the transport design of IRR networks which achieves near-optimal capacity engineering. In particular, a Karmarkar Algorithm optimal solution to the linear programming flow model achieves approximately a 5 to 8 percentage point reduction in network design cost in comparison to the designs of pre-planned dynamic networks solved with heuristic design techniques. The optimization techniques described in this and the previous Journal paper attain significant capital cost reductions and network performance improvements by properly modeling the more efficient operation of IRR networks.     

On three intelligent systems: dynamic neural,fuzzy, and wavelet networks for training trajectory

Intelligent systems cover a wide range of technologies related to hard sciences, such as modeling and control theory, and soft sciences, such as the artificial intelligence (AI). Intelligent systems, including neural networks (NNs), fuzzy logic (FL), and wavelet techniques, utilize the concepts of biological systems and human cognitive capabilities. These three systems have been recognized as a robust and attractive alternative to the some of the classical modeling and control methods. The application of classical NNs, FL, and wavelet technology to dynamic system modeling and control has been constrained by the non-dynamic nature of their popular architectures. The major drawbacks of these architectures are the curse of dimensionality, such as the requirement of too many parameters in NNs, the use of large rule bases in FL, the large number of wavelets, and the long training times, etc. These problems can be overcome with dynamic network structures, referred to as dynamic neural networks (DNNs), dynamic fuzzy networks (DFNs), and dynamic wavelet networks (DWNs), which have unconstrained connectivity and dynamic neural, fuzzy, and wavelet processing units, called neurons, feurons, and wavelons, respectively. The structure of dynamic networks are based on Hopfield networks. Here, we present a comparative study of DNNs, DFNs, and DWNs for non-linear dynamical system modeling. All three dynamic networks have a lag dynamic, an activation function, and interconnection weights. The network weights are adjusted using fast training (optimization) algorithms (quasi-Newton methods). Also, it has been shown that all dynamic networks can be effectively used in non-linear system modeling, and that DWNs result in the best capacity. But all networks have non-linearity properties in non-linear systems. In this study, all dynamic networks are considered as a non-linear optimization with dynamic equality constraints for non-linear system modeling. They encapsulate and generalize the target trajectories. The adjoint theory, whose computational complexity is significantly less than the direct method, has been used in the training of the networks. The updating of weights (identification of network parameters) is based on Broyden–Fletcher–Goldfarb–Shanno method. First, phase portrait examples are given. From this, it has been shown that they have oscillatory and chaotic properties. A dynamical system with discrete events is modeled using the above network structure. There is a localization property at discrete event instants for time and frequency in this example.     

大型挤压油膜阻尼器非线性动力学分析与优化研究进展

挤压油膜阻尼器（Squeeze Film Dampers,SFDs）是旋转机械中常用的一类支承阻尼结构装置,能够改善转子系统的动力特性.当前工程实际中已经大量使用的两类不同结构形式的挤压油膜阻尼器,仍然存在着减振效果不稳定甚至会导致转子失稳,以及阻尼器动力学机制不清楚、建模和分析精度差、设计方法欠缺等理论技术难题.本文首先介绍两类典型SFD的结构形式和主要失效模式,然后详细叙述SFD动力学分析与优化设计方法在发展过程中的代表性研究成果,涉及SFD动力学特性、转子 SFD系统动力学特性研究、SFD动力学设计与优化等几个方面的研究,并对SFD的试验测试技术方面的成果进行评述.在此基础上,探讨目前先进航空发动机用大型挤压油膜阻尼器亟须开展的基础研究任务,特别强调了数据驱动与动力学解析模型融合的SFD动力学建模、分析与设计优化的发展方向.     

一种改进的神经网络非线性预测控制

从建立神经网络非线性预测模型出发,针对BP网络存在收敛速度慢,容易陷入局部最小的缺点,该文在BFGS拟牛顿法的基础上,提出了一种基于并行拟牛顿优化算法的并行拟牛顿神经网络。该并行拟牛顿优化算法采用两个含有不同参数的拟牛顿校正公式,在每次迭代过程中,利用这两个不同的校正公式得到相应的搜索方向,并通过不精确搜索法求取最优步长,最后根据一性能指标取最优的一个搜索方向和相应的步长对网络各层之间的权值进行修正。Matlab仿真结果表明,同BP神经网络和BFGS拟牛顿神经网络相比,该神经网络具有收敛速度快、模型精度高的特点,更适合于实时非线性控制。