基于相对优劣聚类的道路交通安全灰色评价方法

根据我国道路交通安全现状及道路交通安全监督管理的要求,将灰色系统理论应用于道路交通安全评价,为便于各省市间交通安全管理工作比较,提出了相对优劣的灰类评价概念和标准,建立了一套基于相对优劣聚类的交通安全灰色评价体系和模型。利用该模型对2004年我国31个省市道路交通安全相对优劣水平做了测度与分析。     

基于减聚类-自适应神经模糊推理的船舶航向保持控制设计

为解决船舶在非线性和不确定性条件下的常规航向保持控制参数难以确定和性能较差的问题,提出一种基于减法聚类和神经模糊推理系统(SC-ANFIS)的船舶航向保持控制设计。基于鲁棒PID控制,借助减法聚类算法的学习能力对输入样本进行聚类分析,优化模糊量化和模糊规则,继而用神经-模糊推理的方法解决船舶的不确定性问题和非线性控制问题;同时,为避免维数灾难等问题发生,采用多维隶属度函数设计一种可在线自调整的基于SC-ANFIS的航向保持控制系统,并设计仿真试验进行对比分析。仿真试验结果表明,在存在模型参数摄动和干扰的情况下,基于SC-ANFIS的航向保持控制系统可行、有效,能取得良好的控制效果。     

基于有序聚类分析的高速公路路段长度划分研究

提出了基于F isher有序聚类分析的高速公路路段长度划分方法,以事故数沿每公里的空间分布为聚类指标进行路段长度划分,通过实例验证,一方面发现按照F isher有序聚类分析方法划分路段的结果能够使得分段后的事故频数很好地服从一定的概率分布,有益于模型的构建;另一方面,应用该方法分段能够很好地考虑事故的发生水平,把事故水平相近的邻近路段划分到同一研究样本内,有助于准确地量化影响因素的作用,在样本量不是特别大的情况下,能够突出两者之间的关系.     

Probability-based clustering and its application to WLAN location estimation

Wireless local area networks (WLAN) localization based on received signal strength is becoming an important enabler of location based services. Limited efficiency and accuracy are disadvantages to the deterministic location estimation techniques. The probabilistic techniques show their good accuracy but cost more computation overhead. A Gaussian mixture model based on clustering technique was presented to improve location determination efficiency. The proposed clustering algorithm reduces the number of candidate locations from the whole area to a cluster. Within a cluster, an improved nearest neighbor algorithm was used to estimate user location using signal strength from more access points. Experiments show that the location estimation time is greatly decreased while high accuracy can still be achieved.     

A modified genetic algorithm for product family optimization with platform specified by information theoretical approach

Many existing product family design methods assume a given platform, However, it is not an in-tuitive task to select the platform and unique variable within a product family. Meanwhile, most approachesare single-platform methods, in which design variables are either shared across all product variants or not atall. While in multiple-platform design, platform variables can have special value with regard to a subset ofproduct variants within the product family, and offer opportunities for superior overall design. An informationtheoretical approach incorporating fuzzy clustering and Shannon's entropy was proposed for platform variablesselection in multiple-platform product family. A 2-level chromosome genetic algorithm (2LCGA) was proposedand developed for optimizing the corresponding product family in a single stage, simultaneously determiningthe optimal settings for the product platform and unique variables. The single-stage approach can yield im-provements in the overall performance of the product family compared with two-stage approaches, in which thefirst stage involves determining the best settings for the platform and values of unique variables are found foreach product in the second stage. An example of design of a family of universal motors was used to verify theproposed method.     

The analysis of motor vehicle crash clusters using the vector quantization technique

In this paper, a powerful tool for analyzing motor vehicle data based on the vector quantization (VQ) technique is demonstrated. The technique uses an approximation of a probability density function for a stochastic vector without assuming an “a priori” distribution. A self‐organizing map (SOM) is used to transform accident data from an N‐dimensional space into a two‐dimensional plane. The SOM retains all the original data yet provides an effective visual tool for describing patterns such as the frequency at which a particular category of events occurs. This enables new relationships to be identified. Accident data from three cities in Italy (Turin, Milan, and Legnano) are used to illustrate the usefulness of the technique. Crashes are aggregated and clustered crashes by type, severity, and along other dimensions. The paper includes discussion as to how this method can be utilized to further improve safety analysis. Copyright © 2010 John Wiley & Sons, Ltd.     

基于AFC数据的南昌轨道交通车站精细化分类

为满足城市轨道交通车站精细化客运组织需求,对车站进出站客流特性进行有效的分类管理。结合自动售检票系统(auto fare collection,AFC)采集的进出站客流数据,从车站进出站客流总量及时序特性方面入手,提出一种基于K-means算法的双层规划聚类方法对全线所有车站进行聚类并划分车站类型。首先以车站进出站客运总量为特征指标进行上层聚类,得出不同客运规模的车站大类;然后考虑车站进出站客流的时变特征,根据不同时段内的客流变化特点构建特征向量进行下层聚类,识别车站客流的时序分布特性。研究结果表明：利用本算法得到的分类结果与实际高度吻合,不同类别车站在客运规模和时变特性上差异明显。双层K-means聚类分析算法通过把握客运规模和客流时变特征,对车站进行精细划分,为车站的客运组织提供依据。     

基于灰色聚类决策模型的网络防火墙系统安全性评估

针对网络中防火墙系统安全性能评估的问题,提出一种基于灰色聚类决策模型的评估方法,构建了防火墙系统安全性能综合评估模型,通过专家系统评价分值和设置权重,对应用实例进行了评估,实现了对各防火墙系统的综合评估、排序,为防火墙系统安全性能的改进提供了一个较好的评估参照。     

东部温润季冻区公路三级自然区划分析

在全国公路一、二级自然区划基础上,借助模糊数学理论,定量分析处理影响公路自然区划的种种模糊性因素.通过建立模糊聚类分区模型,依据自然因素对公路工程综合影响差异的相近程度对东部温润季冻区进行三级划分,并辅以地理相关分析法对区划界线进行修正.由此得到的区划结果弱化了主观因素影响,更具科学性.同时,应用MapX开发东部温润季冻区公路三级自然区划地理信息系统,建立区划信息空间数据库.运用地理信息系统的空间分析功能,提供高效率的数据管理与信息查询,实现影响公路工程的自然因素信息化和图形化,为各部门的工程技术人员和管理者提供充分的决策依据.     

周期分类和Single-Pass聚类相结合的话题识别与跟踪方法

针对增量式聚类初始时话题模型不够充分和准确,随处理报道数量增加,误检与漏检的累积效应被放大的问题,提出了周期分类和Single—Pass聚类相结合的话题识别与跟踪方法．首先采用增量式聚类算法进行话题识别与跟踪,当新闻文本每积累到一定程度之后,对已经聚类的报道进行周期分类,使话题簇精度提高,从而提高后续话题识别与跟踪精度．实验表明这种方法是有效的,能够降低漏检率与错检率,减少归一化错误识别代价．