Interval standard neural network models for nonlinear systems

INTRODUCTION Neural networks have been successfully em- ployed for controlling nonlinear systems since the 1990’s (Narendra and Parthasarathy 1990; Hunt et al., 1992; Suykens et al., 1996). In these nonlinear control systems, neural networks have been used either for modelling the system to be controlled, or for design- ing a controller, or both. Recently, the robustness issue has been an important focus of research in neuro-control circles (Suykens et al., 1996; Wams et al., 1999; Aya…     

LMI-based approach for global asymptotic stability analysis of continuous BAM neural networks

Studies on the stability of the equilibrium points of continuous bidirectional associative memory (BAM) neural network have yielded many useful results. A novel neural network model called standard neural network model (SNNM) is ad- vanced. By using state affine transformation, the BAM neural networks were converted to SNNMs. Some sufficient conditions for the global asymptotic stability of continuous BAM neural networks were derived from studies on the SNNMs’ stability. These co…     

Robust exponential stability analysis of a larger class of discrete-time recurrent neural networks

The robust exponential stability of a larger class of discrete-time recurrent neural networks (RNNs) is explored in this paper. A novel neural network model, named standard neural network model (SNNM), is introduced to provide a general framework for stability analysis of RNNs. Most of the existing RNNs can be transformed into SNNMs to be analyzed in a unified way. Applying Lyapunov stability theory method and S-Procedure technique, two useful criteria of robust exponential stability for the discrete-time SNNMs are derived. The conditions presented are formulated as linear matrix inequalities (LMIs) to be easily solved using existing efficient convex optimization techniques. An example is presented to demonstrate the transformation procedure and the effectiveness of the results.     

A maximum power point tracker for photovoltaic energy systems based on fuzzy neural networks

To extract the maximum power from a photovoltaic （PV） energy system, the real-time maximum power point （MPP） of the PV array must be tracked closely. The non-linear and time-variant characteristics of the PV array and the non-linear and non-minimum phase characteristics of a boost converter make it difficult to track the MPP for traditional control strategies. We propose a fuzzy neural network controller （FNNC）, which combines the reasoning capability of fuzzy logical systems and the learning capability of neural networks, to track the MPP. With a derived learning algorithm, the parameters of the FNNC are updated adaptively. A gradient estimator based on a radial basis function neural network is developed to provide the reference information to the FNNC. Simulation results show that the proposed control algorithm provides much better tracking performance compared with the filzzy logic control algorithm.     

Stability analysis of discrete-time BAM neural networks based on standard neural network models

To facilitate stability analysis of discrete-time bidirectional associative memory (BAM) neural networks, they were converted into novel neural network models, termed standard neural network models (SNNMs), which interconnect linear dynamic systems and bounded static nonlinear operators. By combining a number of different Lyapunov functionals with S-procedure, some useful criteria of global asymptotic stability and global exponential stability of the equilibrium points of SNNMs were derived. These stability conditions were formulated as linear matrix inequalities (LMIs). So global stability of the discrete-time BAM neural networks could be analyzed by using the stability results of the SNNMs. Compared to the existing stability analysis methods, the proposed approach is easy to implement, less conservative, and is applicable to other recurrent neural networks.     

Nonlinear modelling of a SOFC stack by improved neural networks identification

The solid oxide fuel cell (SOFC) is a nonlinear system that is hard to model by conventional methods. So far,most existing models are based on conversion laws,which are too complicated to be applied to design a control system. To facilitate a valid control strategy design,this paper tries to avoid the internal complexities and presents a modelling study of SOFC per-formance by using a radial basis function (RBF) neural network based on a genetic algorithm (GA). During the process of mod-elling,the GA aims to optimize the parameters of RBF neural networks and the optimum values are regarded as the initial values of the RBF neural network parameters. The validity and accuracy of modelling are tested by simulations,whose results reveal that it is feasible to establish the model of SOFC stack by using RBF neural networks identification based on the GA. Furthermore,it is possible to design an online controller of a SOFC stack based on this GA-RBF neural network identification model.     

Exponential synchronization of general chaotic delayed neural networks via hybrid feedback

This paper investigates the exponential synchronization problem of some chaotic delayed neural networks based on the proposed general neural network model, which is the interconnection of a linear delayed dynamic system and a bounded static nonlinear operator, and covers several well-known neural networks, such as Hopfield neural networks, cellular neural networks （CNNs）, bidirectional associative memory （BAM） networks, recurrent multilayer perceptrons （RMLPs）. By virtue of Lyapunov- Krasovskii stability theory and linear matrix inequality （LMI） technique, some exponential synchronization criteria are derived. Using the drive-response concept, hybrid feedback controllers are designed to synchronize two identical chaotic neural networks based on those synchronization criteria. Finally, detailed comparisons with existing results are made and numerical simulations are carried out to demonstrate the effectiveness of the established synchronization laws.     

新型神经网络的发展及其应用

人工神经网络是指模拟人脑神经系统的结构和功能,运用大量的处理部件,由人工方式建立起来的网络系统。近几年来,人工神经网络的研究工作十分活跃,取得了很大的进展,研究开发出了几十种神经网络的模型,出现了多种新型神经网络结构。本文重点介绍了近年来几种新型神经网络的基本模型及典型应用,包括小波神经网络、模糊神经网络、进化神经网络、细胞神经网络、混沌神经网络,同时比较了这几种新型神经网络的优势和不足。最后,根据这几种新型神经网络的特点,展望了它们今后的发展前景。     

Calculation method of ship collision force on bridge using artificial neural network

Ship collision on bridge is a dynamic process featured by high nonlinearity and instantaneity. Calculating ship-bridge collision force typically involves either the use of design-specification-stipulated equivalent static load, or the use of finite element method （FEM） which is more time-consuming and requires supercomputing resources. In this paper, we proposed an alternative approach that combines FEM with artificial neural network （ANN）. The radial basis function neural network （RBFNN） employed for calculating the impact force in consideration of ship-bridge collision mechanics. With ship velocity and mass as the input vectors and ship collision force as the output vector, the neural networks for different network parameters are trained by the learning samples obtained from finite element simulation results. The error analyses of the learning and testing samples show that the proposed RBFNN is accurate enough to calculate ship-bridge collision force. The input-output relationship obtained by the RBFNN is essentially consistent with the typical empirical formulae. Finally, a special toolbox is developed for calculation effi- ciency in application using MATLAB software.     

机器人中的固有神经网络结构及其应用

文章对存在于机器人中的一类神经网络结构作了详细的分析，由此提出机器人固有神经网络概念，讨论了利用机器人固有神经网络简化控制器设计，实现机器人高速高精度跟踪控制的一种方案。     

Control method for exoskeleton ankle with surface electromyography signals

This paper is concerned with a control method for an exoskeleton ankle with electromyography (EMG) signals. The EMG signals of human ankle and the exoskeleton ankle are introduced. Then a control method is proposed to control the exoskeleton ankle using the EMG signals. The feed-forward neural network model applied here is composed of four layers and uses the back-propagation training algorithm. The output signals from neural network are processed by the wavelet transform. Finally the control orders generated from the output signals are passed to the motor controller and drive the exoskeleton to move. Through experiments, the equality of neural network prediction of ankle movement is evaluated by giving the correlation coefficient. It is shown from the experimental results that the proposed method can accurately control the movement of ankle joint. Project supported by the National High-Tech R&D Program (Grant No.2006AA04Z224), the Innovation Program of Shanghai Municipal Education Commission (Grant No.08ZZ48), and the Shanghai Leading Academic Discipline Project (Grant No.Y0102)     

New hybrid model of proton exchange membrane fuel cell

Model and simulation are good tools for design optimization of fuel cell systems. This paper proposes a new hybrid model of proton exchange membrane fuel cell (PEMFC). The hybrid model includes physical component and black-box com-ponent. The physical component represents the well-known part of PEMFC, while artificial neural network (ANN) component estimates the poorly known part of PEMFC. The ANN model can compensate the performance of the physical model. This hybrid model is implemented on Matlab/Simulink software. The hybrid model shows better accuracy than that of the physical model and ANN model. Simulation results suggest that the hybrid model can be used as a suitable and accurate model for PEMFC.     

基于人工神经网络的等离子熔积层形貌试验研究

1IntroductionPlasma deposition manufacturing(PDM)is a newlydeveloped direct metal fabrication process based onplasma transferred arc surfacing[1],as shown in Fig.1.Unlike most existing rapid prototyping techn-ologies,this technique is characterized by sup…     

Maximum power point tracking of a photovoltaic energy system using neural fuzzy techniques

In order to improve the output efficiency of a photovoltaic (PV) energy system, the real-time maximum power point (MPP) of the PV array should be tracked closely. The non-linear and time-variant characteristics of the photovoltaic array and the non-linear and non-minimum phase characteristics of a boost converter make it difficult to track the MPP as in traditional control strategies. A neural fuzzy controller (NFC) in conjunction with the reasoning capability of fuzzy logical systems and the learning capability of neural networks is proposed to track the MPP in this paper. A gradient estimator based on a radial basis function neural network is developed to provide the reference information to the NFC. With a derived learning algorithm, the parameters of the NFC are updated adaptively. Experimental results show that, compared with the fuzzy logic control algorithm, the proposed control algorithm provides much better tracking performance.     

Parameters optimization and nonlinearity analysis of grating eddy current displacement sensor using neural network and genetic algorithm

A grating eddy current displacement sensor （GECDS） can be used in a watertight electronic transducer to realize long range displacement or position measurement with high accuracy in difficult industry conditions. The parameters optimization of the sensor is essential for economic and efficient production. This paper proposes a method to combine an artificial neural network （ANN） and a genetic algorithm （GA） for the sensor parameters optimization. A neural network model is developed to map the complex relationship between design parameters and the nonlinearity error of the GECDS, and then a GA is used in the optimization process to determine the design parameter values, resulting in a desired minimal nonlinearity error of about 0.11%. The calculated nonlinearity error is 0.25%. These results show that the proposed method performs well for the parameters optimization of the GECDS.     

Adaptive backstepping sliding mode control for nonlinear systems with input saturation

An adaptive backstepping sliding mode control is proposed for a class of uncertain nonlinear systems with input saturation.A command filtered approach is used to prevent input saturation from destroying the adaptive capabilities of neural networks (NNs).The control law and adaptive updating laws of NNs are derived in the sense of Lyapunov function,so the stability can be guaranteed even under the input saturation.The proposed control law is robust against the disturbance,and it can also eliminate the impact of input saturation.Simulation results indicate that the proposed controller has a good performance.     

Hybrid intelligent PID control design for PEMFC anode system

Control design is important for proton exchange membrane fuel cell （PEMFC） generator. This work researched the anode system ofa 60-kW PEMFC generator. Both anode pressure and humidity must be maintained at ideal levels during steady operation. In view of characteristics and requirements of the system, a hybrid intelligent PID controller is designed specifically based on dynamic simulation. A single neuron PI controller is used for anode humidity by adjusting the water injection to the hydrogen cell. Another incremental PID controller, based on the diagonal recurrent neural network （DRNN） dynamic identification, is used to control anode pressure to be more stable and exact by adjusting the hydrogen flow rate. This control strategy can avoid the coupling problem of the PEMFC and achieve a more adaptive ability. Simulation results showed that the control strategy can maintain both anode humidity and pressure at ideal levels regardless of variable load, nonlinear dynamic and coupling characteristics of the system. This work will give some guides for further control design and applications of the total PEMFC generator.     

两自由度闭链机器人的神经网络自适应控制

A closed-chain robot has several advantages over an open-chain robot, such as high mechanical rigidity, high payload, high precision. Accurate trajectory control of a robot is essential in practical use. This paper presents an adaptive proportional integral differential (PID) control algorithm based on radial basis function (RBF) neural network for trajectory tracking of a two-degree-of-freedom (2-DOF) closed-chain robot. In this scheme, an RBF neural network is used to approximate the unknown nonlinear dynamics of the robot, at the same time, the PID parameters can be adjusted online and the high precision can be obtained. Simulation results show that the control algorithm accurately tracks a 2-DOF closed-chain robot trajectories. The results also indicate that the system robustness and tracking performance are superior to the classic PID method.     

基于遗传算法和神经网络的倒立摆控制系统

以单级倒立摆为对象,介绍了一种融合遗传算法的神经网络控制方法。该方法采用以多层前馈神经网络作为遗传搜索表示方法的思想,以神经网络为基础,用遗传算法来学习神经网络的权系数,既保留了遗传算法的强全局随机搜索能力,又具有神经网络的鲁棒性和自学习能力。仿真结果证明:遗传算法和神经网络的结合,可兼有神经网络广泛映射能力和遗传算法快速全局收敛等性能。     

鲁棒自适应飞行跟踪控制系统设计

针对具有外部扰动的一类非线性飞行控制系统提出了一种在线神经网络控制的方法,应用3个在线神经网络分别实时抵消系统中的非线性部分、与控制量耦合的非线性项以及外部扰动,使受控系统的输出可以完全跟踪给定输入参考信号。文中以一个实际的验证机非线性模型为仿真对象,验证了这种方法的有效性,由于在线神经网络的动态补偿特性,即使飞机在不同的高度和速度下,在模型参数发生改变时,控制过程并不需要改变增益表就可以完成全包络的飞行跟踪控制,相关的定理证明了整个神经自适应飞行闭环控制系统的稳定性,仿真示例表明了该设计方法的有效性。