该文提出一种音频情感区间划分方法。该方法以提取音频情感语义方面为目的，可以有效地划分出视频流中音频通道的情感区间。首先，事先选定若干种音频中层情感认知类型，并采用基于分层二叉树SVM分类算法对每个音频段进行中层情感认知初分类，然后提出一种基于规则的分类结果平滑策略对初分类结果进行平滑。最后，利用从中层认知到高层情感感知的映射机制，将中层认知映射到高层情感感知以识别高层情感语义，最终完成音频情感区间划分。实验证明，该方法对音频情感区间划分具有良好的效果。

This paper proposes a emotion perception based division approach of audio emotional range, which starts with the au-dio emotional semantic analysis and works well. Firstly, several kinds of middle-level emotional cognitive type are selected in ad-vance. An hierarchical binary tree based SVM classifier algorithm is performed to classify the middle-level emotional cognitive type initially. Next, for the purpose of finishing emotion labeling, a rule based smoothing strategy for emotional ranges is proposed to smooth the former classification results. Finally, a mapping mechanism, from middle-level cognitive types to high-level emo-tional perceptional types, is adopted to synchronize the high-level audio emotional perception results onto the horizontal axis of audio energy curve. Experimental results demonstrate that the proposed scheme is effective for audio emotional range division.

在模糊时间序列模型的构架中,介绍了广义模糊时间序列模型建立过程和常用的模糊区间划分方法,提出了基于均匀划分、模糊C均值聚类和自动聚类3种模糊区间划分方法的广义模糊时间序列模型,并用Alabama大学入学人数和沪市股指两组数据对模型进行了详细的分析.实验结果不仅揭示了这3种方法对模型预测结果的影响,还证明了广义模型优于传统模型.

In the framework of fuzzy time series model ,the generalized fuzzy time series model and some methods for partitioning fuzzy interval are presented and summarized . Secondly , three generalized fuzzy time series models on the basis of average partition ,fuzzy C-mean (FCM ) clustering and automatic clustering techniques are presented .Enrollment of the University of Alabama and the close prices of Shanghai Stock Exchange Composite Index (SSECI) are served as the training data sets for the proposed models .The empirical analyses not only reveal the impact of three partition methods on the forecast results , but also show that the generalized model outperforms the conventional counterparts .

针对现有连续函数优化蚁群算法对自变量的初始区间存在敏感度问题，提出泛区间搜索的理念。通过在网格策略上加入新元素---自调整定义域的机制、自适应的蚁群规模、自适应的信息素增加强度和自适应的网格划分份数，提出泛区间搜索的连续函数优化蚁群算法。该算法可根据现有区间判断最优解的方位，实现全实数范围内的广度搜索。仿真实验表明该算法具备鲁棒性，在初始区间不含最优解的条件下也能找到最优解，且收敛速度和计算准确性受区间变化的影响较小。

A concept of extensive-domain search is proposed to make ant colony optimization for continuous function overcome the sensitivity to initial domains of independent variables. By adding new elements to the gridding method, such as self-adaptive domain adjustment, self-adaptive ant size, self-adaptive pheromone increment and self-adaptive domain division, extensive-domain-search ant colony optimization (EDS-ACO) is put forward. Thus, the optimal solution can be found by EDS-ACO through an extensive search in the whole range of real numbers. Experiments show that EDS-ACO has the robustness since it can obtain the correct results in the case of initial domains without the optimal solution. The variation of initial domains has a small influence on convergence speed and computational accuracy of EDS-ACO.

提出一种新的有效的FTIR光谱气体浓度反演的方法。该方法将区间划分的思想用于红外光谱波长优化筛选,即将红外光谱在给定波长范围内划分为若干个子区间,在每个子区间中利用遗传算法(genetic algorithm,GA)优化后的极限学习机(extreme learning machine,ELM)建立浓度预测模型,根据每个子区间测试集均方根误差RMSE和相关系数R2的大小评价模型的泛化性能,筛选出最优子区间组合建立预测模型。通过含干扰组分(CO2,N2O)的CO气体的FTIR光谱对提出的算法进行了验证,在波段为2 140~2 220cm-1范围内利用区间法筛选出的最优组合作为变量,应用GA-ELM建立的浓度反演模型,其决定系数R2为0.987 4,均方根误差RMSE为154.996 3,建模时间仅为0.8s,表明该算法(Interval-GA-ELM,iGELM)的应用不仅缩短了建模时间,而且在干扰组分存在的情况下,依然可以准确筛选出特征波长,从而提高了模型稳定性和预测精度,为大气污染气体遥测分析提供了行之有效的方法。

This paper proposed a novel effective quantitative analysis method for FTIR spectroscopy of polluted gases ,which se-lect the best wavenumbers based on the idea of interval dividing .Meanwhile ,genetic algorithm was adopted to optimize the con-nect weights and thresholds of the input layer and the hidden layer of extreme learning machine (ELM ) because of its global search ability .Firstly ,the whole spectrum region was divided into several subintervals ;Secondly ,the quantitative analysis mod-el was established in each subinterval by using optimized GA-ELM ;Thirdly ,the best combination of subintervals was selected according to the generalized performance of each subinterval model by computing the parameters root mean square error (RMSE) and determined coefficients r .In this paper ,the mixture of CO ,CO2 and N2 O gases were selected as the research object and the whole spectrum range was from 2 140 to 2 220 cm -1 .The experiment results showed that the RMSE of model established with

由于传统的生态安全评价方法不能处理区间数和语言值，并且安全度等级划分的主观性和随意性较大，评价结果有效性差。基于二元语义和区间FCM的生态安全评价方法，将实数看成区间数的特例，语言数也转化为区间数，FCM聚类方法将区间值转化为语言值，根据最大隶属度原则，对方案集进行分类。将语言值转化为二元语义，将方案集成结果转换为二元语义，利用二元语义的有序性对方案进行排序，既能处理区间值也能处理语言值，分类标准更加客观。

Because the traditional evaluation method for ecological security can''t deal with interval number and linguistic value, and there are great subjectivity and randomness about the security grade, then the effectiveness of the results is very low. The ecological security evaluation method based on two-tuple linguistic and internal FCM takes the real numbers as a special case of interval number, the number of languages is also converted to interval number, FCM clustering method transforms interval values into linguistic values, according to the principle of maximum degree, the program was set for classification. Converting the linguistic values to two-tuple linguistic, transforming integrated results of the program into two-tuple linguistic and making use of the two-tuple linguistic orderliness sort the program can deal with both the interval value and linguistic value, and classification standard get more objective.