构造性神经网络在测井岩性识别中的应用

测井岩性识别是石油勘探中十分重要的基础工作,准确的岩性识别结果可以为勘探和开发提供可靠的依据。人工神经网络方法可对复杂的高维数据进行非线性映射,在模式识别领域得到越来越广泛的应用。针对传统的BP神经网络算法存在收敛速度慢、隐层数以及隐层节点数难以确定等缺点,研究了一种构造性神经网络学习算法---级联算法（Cascade Correlation Algorithm,简称CC算法）及其在石油工程中的应用。采用该算法对甘肃镇原-泾川地区进行岩性识别研究,通过与BP神经网络的识别结果进行比较,体现出构造性神经网络的优越性。     

Inference and learning in hybrid probabilistic network

This paper proposed a novel hybrid probabilistic network, which is a good tradeoff between the model complexity and learnability in practice. It relaxes the conditional independence assumptions of Naive Bayes while still permitting efficient inference and learning. Experimental studies on a set of natural domains prove its clear advantages with respect to the generalization ability.     

A genetic‐based neuro‐fuzzy controller for blind equalization of time‐varying channels

This paper presents a neuro‐fuzzy network (NFN) where all its parameters can be tuned simultaneously using genetic algorithms (GAs). The approach combines the merits of fuzzy logic theory, neural networks and GAs. The proposed NFN does not require a priori knowledge about the system and eliminates the need for complicated design steps such as manual tuning of input–output membership functions, and selection of fuzzy rule base. Although, only conventional GAs have been used, convergence results are very encouraging. A well‐known numerical example derived from literature is used to evaluate and compare the performance of the network with other equalizing approaches. Simulation results show that the proposed neuro‐fuzzy controller, all parameters of which have been tuned simultaneously using GAs, offers advantages over existing equalizers and has improved performance. From the perspective of application and implementation, this paper is very interesting as it provides a new method for performing blind equalization. The main contribution of this paper is the use of learning algorithms to train a feed‐forward neural network for M‐ary QAM and PSK signals. This paper also provides a platform for researchers of the area for further development. Copyright © 2008 John Wiley & Sons, Ltd.     

基于人工神经网络的混合气体定量检测方法

将气体传感器阵列与人工神经网络相结合，设计了用于混合气体定量检测的人工嗅觉系统；构造了局部连接的前馈神经网络的结构和基于递推预报误差的网络训练算法；通过实验对CO，H2，CH4三种气体的混合物进行了分析，实验结果证明气体传感器阵列和人工神经网络用于混合气体定量检测是可行的。     

On-line emission and economic load dispatch using adaptive Hopfield neural network

This paper presents an adaptive Hopfield neural network (AHNN) based methodology, where the slope of the activation function is adjusted, for finding approximate Pareto solutions for the multi-objective optimization problem of emission and economic load dispatch (EELD). We have placed emphasis on finding solutions quickly, rather than the global Pareto solutions, so that our algorithm can be employed in large on-line power systems where variations in load are quite frequent. To enable faster convergence, adaptive learning rates have been developed by using energy function and applied to the slope adjustment method. The proposed algorithm has been tested on selected IEEE bus benchmark systems. The convergence of AHNN is found to be nearly 50% faster than the non-adaptive version.     

线性分组码的神经网络译码

介绍了神经网络与线性分组码之间的关系，并在文献[1]的基础上证明了软判决译码与求解能量函数最大值之间的等价性，然后以(7，4，3)汉明码为例介绍了神经网络在循环码硬判决、软判决译码中的应用。     

A Bluetooth Routing Protocol Using Evolving Fuzzy Neural Networks

In this paper, a routing protocol which utilizes the characteristics of Bluetooth technology is proposed for Bluetooth-based mobile ad hoc networks. The routing tables are maintained in the master devices and the routing zone radius for each table is adjusted dynamically by using evolving fuzzy neural networks. Observing there exists some useless routing packets which are helpless to build the routing path and increase the network loads in the existing ad hoc routing protocols, we selectively use multiple unicasts or one broadcast when the destination device is out of the routing zone radius coverage of the routing table. The simulation results show that the dynamic adjustment of the routing table size in each master device results in much less reply time of routing request, fewer request packets and useless packets compared with two representative protocols, Zone Routing Protocol and Dynamic Source Routing.     

基于地震属性分析的地层孔隙压力钻前预测模型

钻前准确预测地层孔隙压力是进行井壁稳定状态分析的基础。基于地震和测井信息间的密切联系，提出了利用地震属性预测孔隙压力的新方法。通过神经网络建立已钻井段的井旁地震属性和声波时差测井数据之间的非线性关系模型，以此模型为基础预测未钻地层的声波速度。利用预测结果结合岩石力学模型得到垂直有效应力，最后根据有效应力原理计算地层孔隙压力。该方法有效克服了传统的地震层速度预测模型的不足之处，在油田的实际应用中取得了良好的效果，尤其适用于在勘探新区进行孔隙压力预测。     

自适应RBF神经网络在CDMA移动通信上行链路功率控制中的应用研究

提出了一种自适应RBF神经网络功率控制方案。详细研究了该网络在DS-CDMA通信中，进行上行链路闭环功率控制(基于信扰比(SIR))的应用理论，给出了该网络参数的计算方法。最后用计算机仿真法模拟出该控制器的运行性能。结果表明基于SIR的自适应RBF神经网络功率控制器能自适应地调整移动台的发射功率，使基站接收信号的信扰比始终非常接近于一个常数，且有比定步长功率控制更小的SIR跟踪误差，从而可以降低接收信号的中断概率、提高信道容量。     

用基函数神经网络实现多阈值图象分割

本文介绍了一种用基函数神经网络实现多阈值图象分割的新方法。它从函数逼近的角度研究基于灰度直方图的多阈值分割问题，提出了一种模糊反向传播学习算法，采用该算法的高斯基函数网络能够准确检测直方图中包含的子区域和它们的分布函数，而且速度很快。实验表明本文的方法在实际图象分割中是有效的。