结合数据变换和启发式方法的模糊系统辨识

取代应用原始数据样本构造模糊模型的传统方法，提出应用数据变换技术和启发式方法简化模糊建模过程，对于变换后的数据，首先通过启发式方法确定模糊If-Then规则结合部分非模糊单值（即实数）的初始值，然后通过梯度下降学习方法进行精调。该方法不仅模糊精度较高，而且收敛速度快。仿真实验验证了所提出了优于传统的方法。     

基于模糊QoS满意度的启发式多约束路由算法

QoS路由是实现IP网络服务质量的重要手段.针对一般的QoS路由算法时间复杂度高或者只局限于特定约束的缺点,提出了一种基于模糊QoS满意度的启发式多约束路由算法.首先给出了多约束路由的问题模型和数学描述,然后通过模糊处理各QoS参数的方法构造链路的QoS满意度,在此基础上将QoS满意度与传统最短路径优先相结合,通过启发式搜索快速有效地寻找满足所有约束的路由.仿真结果表明,所提出的路由算法拥有较好的性能.     

适合大规模数据集且基于LLM的0阶TSK模糊分类器

针对传统分类器的泛化性能差、可解释性及学习效率低等问题, 提出0阶TSK-FC模糊分类器.为了将该分类器 应用到大规模数据的分类中, 提出增量式0阶TSK-IFC模糊分类器, 采用增量式模糊聚类算 法(IFCM($c+p$))训练模糊规则参数并通过适当的矩阵变换提升参数学习效率.仿真实验表明, 与FCPM-IRLS模糊分类器、径向基函数神经网 络相比, 所提出的模糊分类器在不同规模数据集中均能保持很好的性能, 且TSK-IFC模糊分类器在大规模数据分类中尤为突出.     

数据驱动的双层次软件过程挖掘方法

为了解决软件过程数据因活动信息及案例属性的缺失而无法应用传统过程挖掘方法的问题,以软件过程数据为研究对象,提出了一种双层次的软件过程挖掘方法.在活动层,提出加权结构连接向量模型对过程日志进行向量化,通过平均活动熵来确定过程日志模糊聚类的结果,将聚类结果作为活动信息支持后续挖掘工作的开展;在过程层,以启发式关系度量为基础,针对非完全循环进行研究,提出了过程层单触发序列循环划分的日志完备性条件,并进一步给出了循环归属的度量方法.基于大量真实软件过程数据的实验结果表明了双层次的软件过程挖掘方法的可行性及正确性.     

一种基于GIS和遗传算法的地域选取方法

地域自动选取是新一代野战决策支持系统中一项不可或缺的关键技术.文中提出了一种基于地理信息系统和遗传算法的地域自动选取的方法.该方法针对数据预处理提出了网格数据模型,针对地形分析提出了一种基于模糊综合评判的地域评估模型,并为了提高效率和加快搜索速度设计和应用了适合地域分析特点的启发式遗传算法.对该方法在炮兵阵地自动选取中实际应用的结果做出了测试和评价,证明了其可行性和有效性.     

一种新的运动模糊图像恢复方法

通过对运动模糊产生原因的分析,提出了一种去运动模糊的新方法。首先应用Hough变换和自相关函数估计出运动模糊的方向和长度,然后应用迭代步长自适应的整体变分模型进行图像恢复。实验结果表明,这样的空间域处理方法,不但可以避免传统的频率域去模糊方法产生的震铃效应,而且该方法具有良好的抗噪性和对运动模糊参数估计误差的低敏感性。     

基于向量识别的启发式路径推测算法

浮动车数据主要是由车辆的轨迹点数据组成,是一种重要的原始数据,可以广泛地用于各种交通应用,如交通管理和控制、路况计算等.但是原始的车辆GPS数据存在定位误差,必须经过路径推测的修正处理才可以应用.传统的路径推测算法主要采用两种方法:渐增式和全局式.两种方法各有优缺点,渐增式方法计算速度快但准确性差,全局式方法准确性好但计算速度慢.通过综合考虑两种传统算法,文中提出了一种基于向量识别的启发式路径推测算法,该算法采用了启发式图搜索方式,导入几何运算的约束条件,根据车辆轨迹点所形成的向量与路网模型比较来进行启发式搜索,并选择车辆所有可能行驶的候选路径.根据全局择优的方式从整体进行比较,确定车辆最有可能的行驶路径.实验结果表明,这种算法能够在复杂路网下,比较准确地推测距离间隔较大的车辆轨迹点,并且能够实时高效地处理大规模数据.     

基于模糊ARMAX模型的模糊建模

提出了一种利用MGS(modified Gram-Schmidt)算法建立模糊ARMAX模型的方法, 给出了基于MGS算法的模型结构和参数辨识的一体化方法. 利用MGS正交变换对通过GK模糊聚类的聚类结果进行变换, 确定对模型贡献大的规则, 删除对模型贡献小的规则, 同时对模型中的参数进行估计. 本文提出的方法能够实现模糊模型的结构和参数的优化. 仿真结果表明, 本文提出的方法能够建立非线性系统的模糊ARMAX模型.     

软件迷惑变换的鲁棒性量化评价

为了方便程序员比较多种迷惑变换方案的优劣,提出了一种量化评价迷惑变换鲁棒性的方法.该方法从软件复杂度变化与代码功能模糊性两个相对独立的层面来刻画迷惑变换导致的鲁棒性.首先,从系统的复杂性与信息的多样性角度建立软件系统复杂度模型,模型包含软件结构、信息流、分支、循环以及元素的嵌套层次,力求从复杂性层面更准确地反映变换对软件的保护;之后,为量化描述迷惑变换的功能模糊度,根据专家指标评分法建立单种迷惑变换模糊度模型,在此基础上建立多种迷惑变换复合模糊度模型;然后,阐述了如何联合所提出的模型实现对单种迷惑变换技术有效性判定与多种迷惑方案的选优,也给出了模型的实现算法及一些示例;最后,通过实例仿真详细展示了模型的工作过程.     

基于分布估计算法的磁浮列车故障综合评判

针对使用传统模糊综合评判方法进行故障级别评判时模型参数难以确定的问题.提出一种基于分布估计算法(EDA)的模糊综合故障评判方法.该方法利用EDA进行模糊评判模型的进化学习,能有效实现模糊模型参数的自动优化,并具有模型易于理解、计算效率高的优点.通过对磁浮列车悬浮系统的仿真实验,结果显示基于分布估计算法的模糊故障综合评判方法能获得优于传统进化算法和其他机器学习方法的评判效果,具有较好的应用价值.     

A novel meta-heuristic based method for deriving priorities from fuzzy pairwise comparison judgments

This paper proposes a new method to derive the priority vector from fuzzy pairwise comparison matrices. Unlike several known methods, the proposed method derives crisp weights from consistent and inconsistent fuzzy comparison matrices. Therefore, the crisp weights obviate the need of additional aggregation and ranking procedures. To derive the priority vector, a Modified Fuzzy Logarithmic Least Square Model (MFLLSM) is proposed. In order to solve the MFLLSM, a framework based on genetic algorithm is proposed. In the proposed framework, a heuristic algorithm of population initialization, a heuristic algorithm for simulating fuzzy numbers and a heuristic algorithm of fitness evaluation are proposed.The solution of the prioritization problem requires finding priorities such that their ratio approximately satisfies the initial judgments. Computational results reveal the superiority of the proposed method in comparison with five well known methods of literature from the viewpoint of satisfaction of initial judgments by the obtained priority vector. It is shown by ten different examples that the deviation of the priorities ratio from initial judgments in the proposed method is less than five existing methods of literature. In addition, unlike several methods of literature, the proposed method considers fuzzy judgments represented by both triangular and trapezoidal fuzzy numbers. Furthermore, the proposed method for the first time considers judgments represented by triangular shaped fuzzy numbers and trapezoidal shaped fuzzy numbers which are discussed in the paper.     

基于模糊邻域粗糙集的启发式属性约简算法

属性约简是机器学习等领域中常用的数据预处理方法。在基于粗糙集理论的属性约简算法中,大多是根据单一的方法来度量属性重要度。为了从多角度对属性达到更为优越的评估效果,首先在已有的模糊邻域粗糙集模型中定义属性依赖度度量,然后根据粒计算理论中知识粒度的概念,在模糊邻域粗糙集模型下提出了模糊邻域粒度度量。由于属性依赖度和知识粒度代表了不同视角的属性评估方法,因此将这两种方法结合起来用于信息系统的属性重要度评估,最后给出一种启发式属性约简算法。实验结果表明,所提出的算法具有较好的属性约简性能。     

基于模糊知识粒度的混合属性约简算法

现实世界中常常包含着海量的、不完整的、模糊及不精确的数据或对象,使得模糊信息粒化成为近年来研究趋势。利用论域上的模糊等价关系定义了模糊粒度世界的模糊知识粒度,给出了新的属性约简条件和核属性计算方法,以便更好地挖掘出潜在的、有利用价值的信息。针对粗糙集在对连续属性约简的过程中容易造成信息缺失和不能对模糊属性处理的现象,提出了一种基于模糊知识粒度对混合决策系统约简的启发式算法,省去了连续属性离散化过程,减少了计算量,为离散值域和混合值域约简提供了统一的方法。最后通过实例验证了其有效性。     

A heuristic error-feedback learning algorithm for fuzzy modeling

Describes a type of fuzzy system with interpolating capability to extract MISO fuzzy rules from input-output sample data through learning. The proposed model inherits many merits from Sugeno-type models and their variations. A heuristic error-feedback learning algorithm associated with the model is suggested. Based on which, the estimator is shown to have a self-adjusting step when approaching a minimum     

Solving fuzzy capacitated location routing problem using hybrid variable neighborhood search and evolutionary local search

A fuzzy capacitated location routing problem (FCLRP) is solved by using a heuristic method that combines variable neighborhood search (VNS) and evolutionary local search (ELS). Demands of the customer and travel times between customers and depots are considered as fuzzy and deterministic variables, respectively in FCLRP. Heterogeneous and homogeneous fleet sizes are performed together to reach the least multi-objective cost in a case study. The multi-objective cost consists of transportation cost, additional cost, vehicle waiting cost and delay cost. A fuzzy chance constrained programming model is added by using credibility theory. The proposed method reaches the solution by performing four stages. In the first stage, initial solutions are obtained by using a greedy heuristic method, and then VNS heuristic, which consists of seven different neighborhood structures, is performed to improve the solution quality in the second stage. In the third stage, a perturbation procedure is applied to the improved solution using ELS algorithm, and then VNS heuristic is applied again in the last stage. The combination of VNS and ELS is called VNSxELS algorithm and applied to a case study, which has fifty-seven customers and five distributing points, effectively in a reasonable time.     

基于一种新模糊模型的非线性系统模糊辨识

提出一种基于新的模糊模型和加权递推最小二乘算法 (WRLSA)的非线性系统模糊辨识方法.新型的具有插值能力的模糊系统可以通过学习从输入输出采样数据中提取MISO系统模糊规则,它继承了Sugeno模型及其变化形式的许多优点.采用相应的模糊隶属函数,使得被辨识的模型可用若干局部线性模型来表示,然后利用WRLSA拟合这些线性模型.给出了详细的模糊辨识算法,为了验证该辨识方法的有效性,还给出了对熟知的Box-Jenkins数据的辨识结果.     

Heuristic constraints enforcement for training of and knowledgeextraction from a fuzzy/neural architecture. I. Foundation

Using fuzzy/neural architectures to extract heuristic information from systems has received increasing attention. A number of fuzzy/neural architectures and knowledge extraction methods have been proposed. Knowledge extraction from systems where the existing knowledge limited is a difficult task. One of the reasons is that there is no ideal rulebase, which can be used to validate the extracted rules. In most of the cases, using output error measures to validate extracted rules is not sufficient as extracted knowledge may not make heuristic sense, even if the output error may meet the specified criteria. The paper proposes a novel method for enforcing heuristic constraints on membership functions for rule extraction from a fuzzy/neural architecture. The proposed method not only ensures that the final membership functions conform to a priori heuristic knowledge, but also reduces the domain of search of the training and improves convergence speed. Although the method is described on a specific fuzzy/neural architecture, it is applicable to other realizations, including adaptive or static fuzzy inference systems. The foundations of the proposed method are given in Part I. The techniques for implementation and integration into the training are given in Part II, together with applications     

Solving fuzzy Multidimensional Multiple-Choice Knapsack Problems: The multi-start Partial Bound Enumeration method versus the efficient epsilon-constraint method

In this paper a new fuzzy Multidimensional Multiple-choice Knapsack Problem (MMKP) is proposed. In the proposed fuzzy MMKP, each item may belong to several groups according to a predefined fuzzy membership value. The total profit and the total cost of the knapsack problem are considered as two conflicting objectives. A mathematical approach and a heuristic algorithm are proposed to solve the fuzzy MMKP. One method is an improved version of a well-known exact multi-objective mathematical programming technique, called the efficient ?-constraint method. The second method is a heuristic algorithm called multi-start Partial-Bound Enumeration (PBE). Both methods are used to comparatively generate a set of non-dominated solutions for the fuzzy MMKP. The performance of the two methods is statistically compared with respect to a set of simulated benchmark cases using different diversity and accuracy metrics.     

FORECASTING FUZZY TIME SERIES ON A HEURISTIC HIGH-ORDER MODEL

ABSTRACT

Chen first proposed the high-order fuzzy-time series model to overcome the drawback of existing fuzzy first-order forecasting models. His model involved easy calculations and forecasted more accurately than the other models. This study proposes an enhanced fuzzy-time series model, called heuristic high-order fuzzy time series model, to deal with forecasting problems. The proposed model aims to overcome the deficiency of Chen's model, which depends strongly on the highest-order fuzzy-time series to eliminate ambiguities at forecasting and requires a vast memory for data storage. The empirical analysis reveals that the proposed model yields more accurate forecasts.     

A Hybrid Learning Method for the Data-Driven Design of Linguistic Dynamic Systems

In lots of data based prediction or modeling applications, uncertainties and/or noises in the observed data cannot be avoided. In such cases, it is more preferable and reasonable to provide linguistic (fuzzy) predicted results described by fuzzy memberships or fuzzy sets instead of the crisp estimates depicted by numbers. Linguistic dynamic system (LDS) provides a powerful tool for yielding linguistic (fuzzy) results. However, it is still difficult to construct LDS models from observed data. To solve this issue, this paper first presents a simplified LDS whose inputoutput mapping can be determined by closed-form formulas. Then, a hybrid learning method is proposed to construct the data-driven LDS model. The proposed hybrid learning method firstly generates fuzzy rules by the subtractive clustering method, then carries out further optimization of centers of the consequent triangular fuzzy sets in the fuzzy rules, and finally adopts multiobjective optimization algorithm to determine the left and right end-points of the consequent triangular fuzzy sets. The proposed approach is successfully applied to three real-world prediction applications which are: prediction of energy consumption of a building, forecasting of the traffic flow, and prediction of the wind speed. Simulation results show that the uncertainties in the data can be effectively captured by the linguistic (fuzzy) estimates. It can also be extended to some other prediction or modeling problems, in which observed data have high levels of uncertainties.