A recurrent network for dynamic system identification

This paper presents a type of recurrent artificial neural network architecture for identification of an arbitrary, continuous dynamic system. The recurrent network is shown to be stable for a constant input with certain conditions on the parameters of the network. The proposed network has significant advantages over similar models in continuous time nonlinear system identification and is used to identify three nonlinear dynamic systems. Finally, the applicability of the radial basis function networks using the same network architecture to reduce the time-complexity of the training task is presented.     

A self‐organizing recurrent fuzzy CMAC model for dynamic system identification

This paper presents a self‐organizing recurrent fuzzy cerebellar model articulation controller (RFCMAC) model for identifying a dynamic system. The recurrent network is embedded in the self‐organizing RFCMAC by adding feedback connections with a receptive field cell to the RFCMAC, where the feedback units act as memory elements. A nonconstant differentiable Gaussian basis function is used to model the hypercube structure and the fuzzy weight. An online learning algorithm is proposed for the automatic construction of the proposed model during the learning procedure. The self‐constructing input space partition is based on the degree measure to appropriately determine various distributions of the input training data. A gradient descent learning algorithm is used to adjust the free parameters. The advantages of the proposed RFCMAC model are summarized as (1) it requires much lower memory requirement than other models; (2) it selects the memory structure parameters automatically; and (3) it has better identification performance than other recurrent networks. © 2008 Wiley Periodicals, Inc.     

A context layered locally recurrent neural network for dynamic system identification

This paper puts forward a novel recurrent neural network (RNN), referred to as the context layered locally recurrent neural network (CLLRNN) for dynamic system identification. The CLLRNN is a dynamic neural network which appears in effective in the input–output identification of both linear and nonlinear dynamic systems. The CLLRNN is composed of one input layer, one or more hidden layers, one output layer, and also one context layer improving the ability of the network to capture the linear characteristics of the system being identified. Dynamic memory is provided by means of feedback connections from nodes in the first hidden layer to nodes in the context layer and in case of being two or more hidden layers, from nodes in a hidden layer to nodes in the preceding hidden layer. In addition to feedback connections, there are self-recurrent connections in all nodes of the context and hidden layers. A dynamic backpropagation algorithm with adaptive learning rate is derived to train the CLLRNN. To demonstrate the superior properties of the proposed architecture, it is applied to identify not only linear but also nonlinear dynamic systems. The efficiency of the proposed architecture is demonstrated by comparing the results to some existing recurrent networks and design configurations. In addition, performance of the CLLRNN is analyzed through an experimental application to a dc motor connected to a load to show practicability and effectiveness of the proposed neural network. Results of the experimental application are presented to make a quantitative comparison with an existing recurrent network in the literature.     

A block-diagonal recurrent fuzzy neural network for system identification

A recurrent fuzzy neural network with internal feedback is suggested in this paper. The network is entitled dynamic block-diagonal fuzzy neural network (DBD-FNN), and constitutes a generalized Takagi-Sugeno-Kang fuzzy system, where the consequent parts of the fuzzy rules are small Block-Diagonal Recurrent Neural Networks. The proposed model is applied to a benchmark identification problem, where a dynamic system is to be identified. Additionally, an application of the proposed model to the problem of the analysis of lung sounds is presented. Particularly, a filter based on the DBD-FNN is developed, trained with the RENNCOM method. Extensive experimental and simulation results are given and performance comparisons with a series of other models are conducted, highlighting the modeling characteristics of DBD-FNN as an identification tool and the effectiveness of the proposed separation filter.     

动态递归神经网络在柴油机系统辨识中的应用

对所提出的动态递归神经网络进行了分析,以及如何利用它们来进行系统辨识.设计了用于辨识柴油机的实验,最后在此基础上对柴油机的模型进行了辨识,取得了较好的效果.     

动态递归神经网络在柴油机系统辨识中的应用

对所提出的动态递归神经网络进行了分析,以及如何利用它们来进行系统辨识。设计了用于辨识柴油机的实验,最后在此基础上对柴油机的模型进行了辨识,取得了较好的效果。     

A new approach to fuzzy-neural system modeling

We develop simple but effective fuzzy-rule based models of complex systems from input-output data. We introduce a simple fuzzy-neural network for modeling systems, and we prove that it can represent any continuous function over a compact set. We introduce “fuzzy curves” and use them to: 1) identify significant input variables, 2) determine model structure, and 3) set the initial weights in the fuzzy-neural network model. Our method for input identification is computationally simple and, since we determine the proper network structure and initial weights in advance, we can train the network rapidly. Viewing the network as a fuzzy model gives insight into the real system, and it provides a method to simplify the neural network     

Nonlinear dynamic system identification using pipelined functional link artificial recurrent neural network

A computationally efficient pipelined functional link artificial recurrent neural network (PFLARNN) is proposed for nonlinear dynamic system identification using a modification real-time recurrent learning (RTRL) algorithm in this paper. In contrast to a feedforward artificial neural network (such as a functional link artificial neural network (FLANN)), the proposed PFLARNN consists of a number of simple small-scale functional link artificial recurrent neural network (FLARNN) modules. Since those modules of PFLARNN can be performed simultaneously in a pipelined parallelism fashion, this would result in a significant improvement in its total computational efficiency. Moreover, nonlinearity of each module is introduced by enhancing the input pattern with nonlinear functional expansion. Therefore, the performance of the proposed filter can be further improved. Computer simulations demonstrate that with proper choice of functional expansion in the PFLARNN, this filter performs better than the FLANN and multilayer perceptron (MLP) for nonlinear dynamic system identification.     

A new pseudo-Gaussian-based recurrent fuzzy CMAC model for dynamic systems processing

This article presents a new pseudo-Gaussian-based recurrent fuzzy cerebellar model articulation controller (PG-RFCMAC) model for identifying various nonlinear dynamic systems. A pseudo-Gaussian basis function can provide the self-organising PG-RFCMAC model, which own a higher flexibility and can approach the optimise result more accurately. The pseudo-Gaussian basis function is used to model the hypercube cells and the fuzzy weights. The recurrent network is embedded in the PG-RFCMAC model by adding feedback connections with a receptive field cell, where the feedback units act as memory elements. An on-line learning algorithm is proposed for the automatic construction of the proposed model during the learning procedure. Computer simulations were conducted to illustrate the performance and applicability of the proposed model.     

Integral-based parameter identification for long-term dynamic verification of a glucose-insulin system model

Hyperglycaemia in critically ill patients increases the risk of further complications and mortality. This paper introduces a model capable of capturing the essential glucose and insulin kinetics in patients from retrospective data gathered in an intensive care unit (ICU). The model uses two time-varying patient specific parameters for glucose effectiveness and insulin sensitivity. The model is mathematically reformulated in terms of integrals to enable a novel method for identification of patient specific parameters. The method was tested on long-term blood glucose recordings from 17 ICU patients, producing 4% average error, which is within the sensor error. One-hour forward predictions of blood glucose data proved acceptable with an error of 2-11%. All identified parameter values were within reported physiological ranges. The parameter identification method is more accurate and significantly faster computationally than commonly used non-linear, non-convex methods. These results verify the model's ability to capture long-term observed glucose-insulin dynamics in hyperglycemic ICU patients, as well as the fitting method developed. Applications of the model and parameter identification method for automated control of blood glucose and medical decision support are discussed.     

A dynamic fuzzy model for a drum-boiler-turbine system

A nonlinear dynamic fuzzy model for natural circulation drum-boiler-turbine is presented. The model is derived from Åström-Bell nonlinear dynamic system and describes the complicated dynamics of the physical plant. It is shown that the dynamic fuzzy model gives in some appropriate sense accurate global nonlinear prediction and at the same time that its local models are close approximations to the local linearizations of the nonlinear dynamic system. This closeness is illustrated by simulation in various conditions.     

An adaptive recurrent fuzzy system for nonlinear identification

This paper describes the architecture and training procedure of a recurrent fuzzy system (RFS). The RFS is composed of a fuzzy inference system (FIS) and a delayed feedback connection. The recurrent property comes from feeding the FIS output back to the FIS input via an adjustable feedback parameter. Both the on-line and off-line training procedures based on the backpropagation-through-time (BPTT) algorithm have been investigated. The adjoint model of the RFS is obtained and used to compute the gradients. It is shown that the off-line training is insufficient to adapt to changes in system dynamics. So, an on-line training procedure is derived. In this procedure, a first in first out stack is used to store a certain history of the input–output data to perform a truncated BPTT algorithm. A quasi-Newton optimization method with a line search algorithm is used to adjust the RFS parameters. The performance of the developed RFS is demonstrated by applying to the identification of nonlinear dynamic systems. The simulation studies show that the proposed identification model has the ability to learn dynamics of highly nonlinear systems and compensate system uncertainties. The results are promising for the further application in the area of control and modeling.     

Deep recurrent Gaussian processes for outlier-robust system identification

Gaussian Processes (GP) comprise a powerful kernel-based machine learning paradigm which has recently attracted the attention of the nonlinear system identification community, specially due to its inherent Bayesian-style treatment of the uncertainty. However, since standard GP models assume a Gaussian distribution for the observation noise, i.e., a Gaussian likelihood, the learning and predictive capabilities of such models can be severely degraded when outliers are present in the data. In this paper, motivated by our previous work on GP learning with data containing outliers and recent advances in hierarchical (deep GPs) and recurrent GP (RGP) approaches, we introduce an outlier-robust recurrent GP model, the RGP-t. Our approach explicitly models the observation layer, which includes a heavy-tailed Student-t likelihood, and allows for a hierarchy of multiple transition layers to learn the system dynamics directly from estimation data contaminated by outliers. In addition, we modify the original variational framework of standard RGP in order to perform inference with the new RGP-t model. The proposed approach is comprehensively evaluated using six artificial benchmarks, within several outlier contamination levels, and two datasets related to process industry systems (pH neutralization and heat exchanger), whose estimation data undergo large contamination rates. The simulation results obtained by the RGP-t model indicates an impressive resilience to outliers and a superior capability to learn nonlinear dynamics directly from highly outlier-contaminated data in comparison to existing GP models.     

A self-adaptive agent-based fuzzy-neural scheduling system for a wafer fabrication factory

A self-adaptive agent-based fuzzy-neural system is constructed in this study to enhance the performance of scheduling jobs in a wafer fabrication factory. The system integrates dispatching, performance evaluation and reporting, and scheduling policy optimization. Unlike in the past studies a single pre-determined scheduling algorithm is used for all agents, in this study every agent develops and modifies its own scheduling algorithm to adapt it to the local conditions. To stabilize the performance of the self-adaptive agent-based fuzzy-neural scheduling system, some treatments have also been taken. To evaluate the effectiveness of the proposed methodology and to make comparison with some existing approaches, production simulation is also applied in this study to generate some test data. According to experimental results, the self-adaptive agent-based fuzzy-neural system did improve the performance of scheduling jobs in the simulated wafer fabrication factory, especially with respect to the average cycle time and cycle time standard deviation.     

Non-linear dynamic system identification: A multi-model approach

We are concerned with models which are able to describe multiple-input multiple-output (MIMO) non-linear dynamic systems. These models are represented in the form of rules and are known as Tagaki-Sugeno models. An identification algorithm for these models based on input and output data is presented. Parameter estimation is based on the calculation of model sensitivity functions with respect to their parameters. Some aspects of structure identification are also tackled, i.e. determination of local model orders and number of rules.     

Nonlinear dynamic system identification for automotive crash using optimization: A review

Literature on linear and nonlinear dynamic system identification is reviewed. The main motivation is to document the state-of-the-art, allowing one to propose further advancements in this field. The main problem is to identify system properties when experimental/numerical input and output data are specified. Parametric as well as nonparametric approaches for system identification are reviewed. For linear systems, both the first order and second order forms of the equations of motion are discussed. The use of first order form is more general as it can treat nonproportional structural damping as well. For nonlinear systems, the second order form of the equations of motion is used. A conclusion from the study is that more work is needed to develop identification formulations for nonlinear dissipative dynamic systems, especially for multi-degree of freedom systems. Received September 11, 2001     

New dynamic fuzzy structure and dynamic system identification

In this study, a new fuzzy system structure that reduces the number of inputs is proposed for dynamic system identification applications. Algebraic fuzzy systems have some disadvantages due to many inputs. As the number of inputs increase, the number of parameters in the training process increase and hence the classical fuzzy system becomes more complex. In the conventional fuzzy system structure, the past information of both inputs and outputs are also regarded as inputs for dynamic systems, therefore the number of inputs may not be manageable even for single input and single output systems. The new dynamic fuzzy system module (DFM) proposed here has only a single input and a single output. We have carried out identification simulations to test the proposed approach and shown that the DFM can successfully identify non-linear dynamic systems and it performs better than the classical fuzzy system.     

交通流模型参数的近似动态规划辨识方法

针对交通流模型的强非线性、不确定性等特点,提出了基于近似动态规划的交通流模型参数辨识算法.该算法具有自学习和自适应的特性,不依赖于被控对象的解析模型.严格的理论推导证明了这种参数辨识方案的收敛性,仿真结果验证了所提出算法的有效性.     

Design of input sequence for linear dynamic system identification

The design of stochastic input sequences for parameter identification of linear dynamic discrete-time system is considered. The determinant of the normalized error covariance matrix is used as a criterion for parameter identification quality which is, as it is shown, dependent on the input autocorrelation function. For a given form of this function, one may derive the optimal values of its parameters. An analytic example illustrating the method is given.     

动态称重系统的建模及其参数估计

阐述了为了兼顾动态称重系统的快速性和精度,将动态称重作为一个基于最小二乘法的参数估计和预测问题来处理,即从建立数学模型和信号处理算法方面加以解决。然后,将动态称重系统等效为二阶系统,分析得出了系统为时变非线性系统,推导出了系统的动态数学模型,并且,根据系统模型,将问题转化为参数辨识问题。辨识算法上,采用了基于Householder变换的自适应最小二乘法,其具有抗方程病态性好、稳定性好、估计精度高、计算量小、跟踪性好等优点。试验结果证明:所提出方法是可行的,达到了试验提出的技术要求,测量相对误差小于0.25%FS,系统在全量程范围内的准确度不低于0.25%。在提高称重速度的同时,也保证了系统的测量准确性,对于此类系统的实用化开发具有很好的参考价值。