基于主动声纳的水下目标特征提取技术综述

主要针对目前国内外常用的主动声纳水下目标特征提取技术进行综述，并分析了目前尚存在的主要技术困难。     

矸石山风蚀扬尘起尘风速确定研究

利用实测空气颗粒物数据和颗粒物体源扩散模式推导扬尘实际排放强度,根据扬尘实际排放强度突变比例确定不同地区矸石山风蚀扬尘起尘风速。     

井下绳类落物打捞工艺技术应用现状分析

井下绳类落物打捞工艺技术是石油井下工程事故处理工艺的重要组成部分。由于绳类落物具有细长、柔软、易滑脱、容易收缩变形的特点,因此在打捞中不易捞获,看似简单,实际打捞还有相当的难度,还有一定的技巧。本文就绳类落物打捞工艺技术进行探讨分析。     

Small Angle Neutron Scattering in P+-Doped Porous Silicon

On the surface of single crystal silicon wafers, porous layers can be formed by electrochemical etching and their structural properties are determined by the doping type and concentration of the substrate. In p⁺-type doped, (1 0 0) oriented wafers the porous structure consists of tube-like voids and column-like remains of the silicon matrix, all perpendicular with respect to the wafer surface. In small angle neutron scattering experiments the micrometer long and nanometer diameter elongated scattering elements, i.e., tubes and columns can be well represented and approximated by cylindrical form factors. In an earlier experiment the diameter and scattering length distribution of the cylindrical scattering elements were measured in a p⁺-type (1 0 0) oriented porous silicon wafer [G. Kádár, G. Káli, Cs. Dücsõ, and E.B. Vázsonyi, Physica B 234–236, 1014 (1997)] and the tube diameters were seen to vary in the range from about 10 to 24 nm. In this paper the continuing small angle neutron scattering study of porous silicon layers will be presented. The evaluation conditions and method for the measured neutron intensity distributions will be discussed. The pore diameter distribution data calculated from the neutron scattering intensity curves are collected and compared in various samples of (1 0 0) oriented p⁺-doped wafers prepared with different porosity and different layer depth. The structural results and data obtained by small angle neutron scattering experiments may help in understanding the practically useful chemical, electronic and other properties of porous silicon.     

基于遗传算法和神经网络的复杂对象的建模与优化

针对复杂对象全局优化结果较多,难以选取最佳参数组合的问题,提出了一种新适应度函数,其特点是使用了目标函数关于其影响参数的梯度之和。结合提出的新适应度函数,首先进行已知模型的遗传算法寻优,然后对神经网络构建的模型进行遗传算法寻优。两次试验结果都表明,搜索到的最佳性能目标对应的参数组合具有较强的鲁棒性。     

Perceptually Grounded Language Learning: Part 2—DETE: A Neural/Procedural Model

In Part 1 of this two-part series, we introduced Katamic memory—a neural network architecture capable of robust sequence learning and recognition. In Part 2, we introduce the Blobs World taskjdomain for language learning and describe the DETE language learning system, which is composed of over 50 Katamic memory modules. DETE currently learns small subsets of English and Spanish via association with perceptual! motor inputs. In addition to Kaiamic memory, DETE employs several other novel features: (1) use of feature planes, to encode visual shapes, spatial relationships and the motions of objects, (2) phase-locking of neural firing, in order to represent focus of atention and to bind objects across multiple feature planes, and (3) a method for encoding temporal relationships, so that DETE can learn utterances involving the immediate past and future. We compare DETE to related models and discuss the implications of this approach for language-learning research.     

基于有限状态机的煤矿辅助运输转载自动化控制研究

为实现辅助运输过程中转载容器装载和转载工作的少人化、智能化,提出一种采用煤矿辅助运输转载机器人实现容器自动化转载的控制方案。概述煤矿辅助运输转载机器人组成及工作原理,分析机器人转载过程,确定了辅助运输转载自动化需要对单轨吊车、转载机械臂、平板车、锁具的工作时序进行控制,在MATLAB中建立系统状态机,将单轨吊车、转载机械臂、平板车、锁具作为子系统,系统状态机根据用户指令结合当前各子系统的状态,协同控制各个子系统的工作时间和顺序,保证了煤矿辅助运输转载的安全进行;最后在ROS中建立辅助运输转载仿真环境,通过系统状态机控制机械臂进行多次抓取实验,实验结果表明转载容器在20次不同位置抓取实验中,抓取成功率均在80%以上。     

基于Canny边缘检测算子和去除小物体算法的熔池图像处理

采用基于Canny边缘检测算子和形态学去除小物体算法提出了一种新的熔池图像处理算法．首先对熔池的图像所具有的一些特性进行分析，然后对图像分别做中值滤波，二值化，边缘检测，和去除小物体处理。实验结果表明，该方法能够比较有效地得出熔池图像的边缘。     

电火花线切割加工中实体数据格式的程序转换

针对电火花线切割编程系统绘制、编辑复杂曲线图形、文字功能差以及数据文件兼容功能不强的弱点，本文提出一种综合解决方案。利用AutoCAD强大的图形能力，基于对其图形实体数据表的分析，对AutoLISP程序提取实体数据，并自动转换成线切割自动编程系统所需的数据文件。该方法可使电火花线切割系统完成复杂图形、字体及具有尖角造型实体的加工，实践证明有较大的实用意义。     

一种改进的CURE聚类算法

聚类分析是数据挖掘领域的一个重要研究方向。已经有多种用于大规模数据库的聚类算法,CURE就是一个典型的代表。本文对CURE进行了改进,新方法用多点表示一个类,但舍弃了代表点收缩的过程;通过对类内最邻近距离统计特征的分析,提出了自动分离子类的方法,因而不用预先给定聚类个数;在CURE对原始数据进行随机采样和分区聚类的基础上,增加了划分网格一步,能降低噪声影响并缩短聚类时间。对二维数据的测试表明：改进的CURE能正确识别大多数类,速度上优于原算法。