A maximum neural network algorithm for route selection problems in multihop radio networks

In a multihop radio network, packets are transmitted from course nodes to destination nodes by activating several links between nodes. Each node can either send a packet to, or receive a packet from, at most one of its adjacent nodes simultaneously. To minimize the transmission time for given requests, the problems must be solved by selecting a transmission route for each request (the routing problem) and by finding a link activation schedule (the link activation problem). The routing problem is decomposed into two subproblems: the candidate extraction problem and the route selection problem. In this paper, we propose a neural network algorithm using the maximum neuron model for the route selection problem. We verify through simulations that our algorithm finds better solutions in a shorter time than the existing algorithms. We also probe the NP‐hardness of this problem. © 1999 Scripta Technica, Electr Eng Jpn, 129(2): 78–86, 1999     

Pseudomemories of two‐dimensional multistate hopfield neural networks

Complex‐valued Hopfield neural networks (CVHNNs) are available for storage of multilevel data, such as gray‐scale images. Such networks have low noise tolerance. This is a severe problem for their applications. To improve the noise tolerance, we have to study pseudomemories. In the case of one training pattern, CVHNNs have only rotated patterns as pseudomemories. There are many rotated patterns. This is considered the reason why CVHNNs have low noise tolerance. In the present paper, we investigate the pseudomemories of two‐dimensional multistate Hopfield neural networks, including the complex‐valued ones, with multiple training patterns. Computer simulations show that there are many pseudomemories other than the rotated patterns. © 2016 Institute of Electrical Engineers of Japan. Published by John Wiley & Sons, Inc.     

An improved elman neural network controller based on quasi‐ARX neural network for nonlinear systems

An improved Elman neural network (IENN) controller with particle swarm optimization (PSO) is presented for nonlinear systems. The proposed controller is composed of a quasi‐ARX neural network (QARXNN) prediction model and a switching mechanism. The switching mechanism is used to guarantee that the prediction model works well. The primary controller is designed based on IENN using the backpropagation (BP) learning algorithm with PSO. PSO is used to adjust the learning rates in the BP process for improving the learning capability. The adaptive learning rates of the controller are investigated via the Lyapunov stability theorem. The proposed controller performance is verified through numerical simulation. The method is compared with the fuzzy switching and 0/1 switching methods to show its effectiveness in terms of stability, accuracy, and robustness. © 2014 Institute of Electrical Engineers of Japan. Published by John Wiley & Sons, Inc.     

Accurate one‐terminal fault location algorithm based on the principle of short‐circuit calculation

An alternative method to find the line fault distance in a transmission network employing only one‐terminal measured data is presented. The proposed method applies Z_bus for short‐circuit calculation to find the fault location on a transmission line without the necessity to know the fault type a priori. The well‐known drawback of the standard simple‐reactance one‐terminal algorithm, which neglects the effect of fault impedance, will be minimized by estimating the voltage drop at the fault location by employing the Z_bus technique. Accuracy the proposed method is demonstrated using the short‐circuit simulation of the modified IEEE‐14 bus test system on MATLAB/Simulink and the Simpower Toolbox. Compared to the accuracy obtained from the standard one‐terminal algorithm, test results confirm substantially improved accuracy of the proposed method in all cases of the four types of fault categories: single line‐to‐ground fault; double line‐to‐ground fault; line‐to‐line fault; and balance three‐phase fault. While the accuracy has been significantly improved, especially for the case with a relatively high fault impedance, also the simplicity in the involved computation is well preserved when compared to other iterative‐based techniques. © 2012 Institute of Electrical Engineers of Japan. Published by John Wiley & Sons, Inc.     

基于BP神经网络的光伏阵列故障诊断研究

光伏阵列多安装在较恶劣的室外环境中,因此在运行过程中常会发生故障。为辨别光伏阵列故障类型,提出了基于L-M算法的BP神经网络的故障诊断方法。在深入分析不同故障状态下光伏阵列输出量变化规律的基础上,确定了故障诊断模型的输入变量。本方法无需额外的设备支持,具有简便、成本低的优点;可以在线实时地进行故障诊断。仿真和初步实验结果验证了基于BP神经网络的故障诊断方法可以有效地检测出光伏阵列短路、断路、异常老化及局部阴影等四种故障。     

A faster method for exactly evaluating one‐to‐all reliabilities of telecommunications networks

A previously proposed method for evaluating one‐to‐all reliability quickly is only an approximation, and its error estimation is problematic because the error bounds are estimated after the approximation algorithm is executed. Many trial‐and‐error executions are required to reach sufficient error bounds. A method for exact evaluation has now been developed that does not require trial‐and‐error executions to determine whether the evaluation error is sufficiently small. The key idea is the application of a factoring method to evaluate one‐to‐all reliability. Evaluation examples show that this method achieves sufficiently quick evaluation that is clearly faster than the previously proposed method. This method will thus be useful for designing telecommunications networks in developing countries. © 2017 Institute of Electrical Engineers of Japan. Published by John Wiley & Sons, Inc.     

基于蚁群算法的神经网络配电网故障选线方法

为了克服基于神经网络的故障选线方法收敛速度慢、易于陷入局部极小点的缺点,提出了蚁群算法和神经网络相结合的故障选线方法。利用ATP-EMTP做单相接地仿真试验,得到各线路的零序电流信号,通过小波变换和傅里叶变换提取其中的故障特征作为神经网络的输入。利用蚁群算法对神经网络进行训练,完成训练的神经网络模型即可实现故障选线。仿真结果表明,该方法训练速度快、误判率低。     

Interlaced switch boxes placement for three‐dimensional FPGA architecture design

Three‐dimensional (3D) field programmable gate array (FPGA) has evoked significant interest in wire‐length reduction for routing requirement. However, the complex design of the 3D switch boxes will limit the performance improvement and suffer from the area efficiency problems. This paper proposed a systematic graph model (SGM) for 3D switch boxes design to simplify the design process and reduce the storage memory for path programming. An interlaced 3D switch boxes and two‐dimensional (2D) switch boxes placement topology is also presented in this paper to design the 3D FPGA architecture for area efficiency purpose. The 3D place and route tool and TSMC 0.18‐µm CMOS process parameters are used to support building the experimental flow for verification. Performance evaluation shows that about 50% storage memory reduction can be obtained by using the proposed SGM‐based switch design approach. Additionally, compared with conventional architectures of 2D FPGA, the proposed scheme based on interlaced switch boxes placement approach can approximately achieve 20% delay‐power product improvement and 43% area‐delay product reduction. Copyright © 2010 John Wiley & Sons, Ltd.     

Implementation of the symmetric doped double‐gate MOSFET model in Verilog‐A for circuit simulation

Recently we developed a model for symmetric double‐gate MOSFETs (SDDGM) that, for the first time, considers the doping concentration in the Si film in the complete range from 1×10¹⁴ to 3×10¹⁸ cm⁻³. The model covers a wide range of technological parameters and includes short channel effects. It was validated for different devices using data from simulations, as well as measured in real devices. In this paper, we present the implementation in Verilog‐A code of this model, which allows its introduction in commercial simulators. The Verilog‐A implementation was optimized to achieve reduction in computational time, as well as good accuracy. Results are compared with data from 2D simulations, showing a very good agreement in all transistor operation regions. Copyright © 2009 John Wiley & Sons, Ltd.     

基于遗传算法的BP神经网络电子系统状态预测方法研究

BP神经网络是一种应用面较广的神经网络,但存在明显缺陷:学习收敛速度慢,易陷入局部极小。遗传算法具有良好的搜索全局最优解的能力。为了提高BP神经网络预测模型在状态预测中的准确性,提出了一种基于遗传算法优化BP神经网络的状态预测方法.利用遗传算法优化BP神经网络的权值和阈值,然后训练BP神经网络预测模型以求得最优解,并将该预测方法应用到Buck输出电压平均值进行有效性验证。仿真结果表明,改进后方法具有较好的非线性拟合能力和更高的预测准确性。     

Decentralized adaptive multi‐dimensional Taylor network tracking control for a class of large‐scale stochastic nonlinear systems

This paper investigates the problem of adaptive multi‐dimensional Taylor network (MTN) decentralized tracking control for large‐scale stochastic nonlinear systems. Minimizing the influence of randomness and complex nonlinearity, which increases computational complexity, and improving the controller's real‐time performance for the stochastic nonlinear system are of great significance. With combining adaptive backstepping with dynamic surface control, a decentralized adaptive MTN tracking control approach is developed. In the controller design, MTNs are used to approximate nonlinearities, the backstepping technique is employed to construct the decentralized adaptive MTN controller, and the dynamic surface control technique is adopted to avoid the “explosion of computational complexity” in the backstepping design. It is proven that all the signals in the closed‐loop system remain bounded in probability, and the tracking errors converge to a small residual set around the origin in the sense of a mean quartic value. As the MTN contains only addition and multiplication, the proposed control method is more simplified and of good real‐time performance, compared with the existing control methods for large‐scale stochastic nonlinear systems. Finally, a numerical example is presented to illustrate the effectiveness of the proposed design approach, and simulation results demonstrate that the method presented in this paper has good real‐time performance and control quality, and the dynamic performance of the closed‐loop system is satisfactory.     

Quasi‐ARX neural network based adaptive predictive control for nonlinear systems

In this paper, a new switching mechanism is proposed based on the state of dynamic tracking error so that more information will be provided –not only the error but also a one up to pth differential error will be available as the switching variable. The switching index is based on the Lyapunov stability theory. Thus the switching mechanism can work more effectively and efficiently. A simplified quasi‐ARX neural‐network (QARXNN) model presented by a state‐dependent parameter estimation (SDPE) is used to derive the controller formulation to deal with its computational complexity. The switching works inside the model by utilizing the linear and nonlinear parts of an SDPE. First, a QARXNN is used as an estimator to estimate an SDPE. Second, by using SDPE, the state of dynamic tracking error is calculated to derive the switching index. Additionally, the switching formula can use an SDPE as the switching variable more easily. Finally, numerical simulations reveal that the proposed control gives satisfactory tracking and disturbance‐rejection performances. Experimental results demonstrate its effectiveness. © 2015 Institute of Electrical Engineers of Japan. Published by John Wiley & Sons, Inc.     

Electromagnetic non‐destructive evaluation by vector finite‐element neural networks

This study proposes neural network‐based iterative inverse solutions for non‐destructive evaluation (NDE) in which vector finite elements (VFEM) represent the forward model that closely models the physical process. The iterative algorithm can eventually estimate the material parameters. Vector finite element method global matrix is stored in a compact form using its sparsity and symmetry. The stored matrix elements are employed as the neurons weights, and preconditioning techniques are used to accelerate convergence of the neural networks (NN) algorithm. Detailed algorithm describing this new method is given to facilitate implementation. Combining vector finite elements and NNs offers several advantages over each technique alone, such as reducing memory storage requirements and the easily computed fixed weights of the NN. Various examples are solved to show the performance and usefulness of the proposed method, including lossy printed circuit board and lossy inhomogeneous cylindrical problems with ferromagnetic materials. These solutions compare very well with other published data where the maximum relative error was 5%. Copyright © 2010 John Wiley & Sons, Ltd.     

蚁群算法优化RBF神经网络的光伏发电MPPT

针对如何提高光伏电池最大功率点预测跟踪精度的问题,讨论了光伏电池非线性输出特性。在此基础上,建立基于RBF神经网络的光伏电池最大功率点预测跟踪模型,以光照强度和温度为模型的输入,光伏电池的最大功率点电压为模型输出,并将RBF神经网络的中心参数采用蚁群算法进行优化,从而实现最大功率点预测跟踪。仿真结果表明,该方法比传统的RBF神经网络方法具有更高的预测精度,能更好地预测跟踪光伏电池的最大功率点。     

A one‐converter contactless charger for electric vehicles

A simple and reliable contactless battery charger for electric vehicles is proposed. Its feature is a unitized power factor correcting (PFC) converter and a high‐frequency inverter (HFI) where the low‐side switch of the HFI also works as the boost switch of the PFC in discontinuous conduction mode, which results in a high input power factor and low harmonic distortion without any feedforward control. The exiting current of the inductive connector, compliant with SAEJ1773, works effectively to make the converter operate in zero voltage switching (ZVS) condition. Another feature is that the charger is controlled by a single magnetically coupled variable frequency oscillator developed by the authors. This paper analyzes the circuits, gives a design example, describes the inductive connector and the oscillator, and presents experimental results. A 1.7‐kW output prototype charger achieved a charging efficiency of 87.4% for total one cycle charging, an overall efficiency of 90% at heavy load, and an input power factor of over 98%. © 2000 Scripta Technica, Electr Eng Jpn, 132(2): 73–81, 2000     

Output‐feedback adaptive tracking control of stochastic nonlinear systems using multi‐dimensional Taylor network

In this paper, an adaptive multi‐dimensional Taylor network (MTN) control scheme based on the backstepping and dynamic surface control (DSC) is developed to solve the tracking control problem for the stochastic nonlinear system with immeasurable states. The MTNs are used to approximate the unknown nonlinearities, and then based on the multivariable analog of circle criterion, an observer is first introduced to estimate the immeasurable states. By combining the adaptive backstepping technique and the DSC technique, an adaptive MTN output‐feedback backstepping DSC approach is developed. It is shown that the proposed controller ensures that all signals of the closed‐loop system are remain bounded in probability, and the tracking error converges to an arbitrarily small neighborhood around the origin in the sense of probability. Finally, the effectiveness of the design approach is illustrated by simulation results.     

基于遗传神经网的自适应电池荷电态预估模型

电池荷电态(SOC)是放电电流、端电压、温度等多种因素的复杂的非线性函数,而且不同类型的电池具有很大的差异,不能建立统一的模型。因此要对其做出精确的预估是一件很困难的事情,需要耗费很多的人力和时间对特定类型的电池进行大量试验然后建模。为克服这些缺点,提出一种基于遗传神经网的自适应SOC预估模型,通过遗传算法对神经网络结构及其学习算法进行优化,在较短的时间内寻找到适合特定类型电池的神经网络模型,大大缩短了人工建模需要的时间,提高了模型对SOC预估的性能。对于三种不同类型电池的数据进行建模的仿真试验结果验证了本方法的有效性。     

一种基于FPGA并行加速的稀疏矩阵求解方法

研究了电力系统电磁暂态仿真中最耗时的稀疏矩阵快速求解问题.采用了算法定义架构的设计思想,提出了一种DAG静态并行调度算法,并设计了与之相适配的硬件并行加速阵列架构.在设计实现中,针对电磁暂态仿真运算中稀疏矩阵求解的特性,采用了精确的节拍级硬件资源调度,实现了高度融合的软硬件协同加速.在此基础上,进行了该设计的测试及性能...     

基于神经网络的CT脑血管图像边缘检测算法

CT脑血管医学图像的三维重构都是源自二维断层扫描,脑血管边缘特征向量的提取是图像处理的关键步骤。为提高边缘特征的提取和保证三维重建图像的质量,在分析了某些常用的边缘检测算法性能基础上,同时结合CT脑血管图像的像素结构特点,将SA_SOFM神经网络算法成功地用于对CT脑血管图像的边缘特征信息提取中。并对算法进行有效的改进,基于真实图像的实验表明该算法提高了边缘特征信息的精度和鲁棒性。     

基于WNN的电阻层析成像图像重建算法的研究

本文对于电阻层析成像系统的图像重建提出了一种基于小波神经网络的图像重建算法,介绍了电阻层析成像技术的原理、数学模型及问题描述,给出了小波神经网络的结构、特点及图像重建的过程.利用MATLAB编程进行了图像重建的仿真实验.仿真结果表明,与线性反投影算法相比,小波神经网络算法提高了重建图像的质量.用横截面图像误差作为重建图像质量的评判标准,线性反投影算法的误差超过了10%,而小波神经网络的算法低于5%.