基于Chebyshev多项式的神经网络中长期负荷预测研究

高精度负荷预测在提高电力系统的安全性和经济性方面有着极其重要的意义,而现有的负荷预测方法因参数有限,难以完全反映其内在规律,因而导致预测结果不够准确.为此提出了一种基于Chebyshev多项式神经网络模型的预测方法.该方法使用递推最小二乘法训练神经网络权值系数,以获得高精度的参数估计,从而实现Chebyshev多项式神经网络模型对负荷量的最优拟合,再利用训练好的Chebyshev多项式神经网络模型实现中长期负荷预测.研究结果表明,该方法能较好模拟负荷变化规律,有效提高了负荷预测精度,在电力系统负荷预测中有较大的应用价值.     

基于多项式神经网络模型的人口预测方法研究

为了提高人口预测精度,提出了基于多项式神经网络模型与递推最小二乘法的人口预测方法.方法完全避免了人为假设条件,充分利用我国六次人口普查数据来建立基于多项式神经网络模型的人口预测模型,并使用递推最小二乘算法递推计算多项式神经网络模型的加权系数.方法能有效预测中长期人口数据及其变化趋势.研究结果表明,中国将在2016年达到人口高峰1385亿.     

灰色预测模型算法的改进研究

为了解决固定模型预测时变系统容易出现较大误差的问题,提出模型更新算法,即采用将移动窗算法与传统灰色预测模型相结合的方法.通过在建模序列中删除一部分旧数据、纳入一部分新数据的方式递推更新预测模型,并分解数学模型所涉及的关键量ab从而简化递推数学公式;利用国家统计年鉴的统计数据验证上述方法的有效性.为了解决传统灰色预测模型建模长度选择的问题,而递推算法能在已知模型参数基础上通过简单计算获得新模型的各项参数,文章给出了详细的递推数学公式;另外对于最小建模长度L_(min)也进行讨论,认为L_(min)≥3,并给出证明.所改进算法提高灰色模型的预测精度,同时也为最优序列长度计算提供理论依据.     

树结构变窗宽局部多项式拟合方法

变窗宽局部多项式模型是探索被估计曲线复杂变化结构的有力工具,其关键思想在于针对不同变化模式的数据区间选择不同窗宽参数进行拟合.基于回归树模型提出一种变窗宽多项式拟合方法,利用回归树的分类功能识别具有不同变化结构的数据区间,在每个子区间上独立选取最优窗宽,通过局部多项式回归实现复杂结构曲线的变窗宽拟合.随机模拟的结果表明,回归树能够有效识别具有不同变化模式的数据区间,变窗宽局部多项式拟合均方误差小,具有计算高效、结果易解释的特点.     

基于最小二乘法的改进GM(1,1)模型在上海市社会总抚养比预测中的应用

由于上海市社会总抚养比受到诸多因素的影响,导致数据波动性较大,如果采用普通灰色预测模型无法更加准确地进行预测,因此文章提出了一种改进的GM(1,1)模型.通过将改进模型与普通模型进行对比得出采用二次多项式拟合的改进模型预测精度较高,因此文章采用二次多项式模拟的改进模型进行预测.将该改进模型应用于上海市社会总抚养比预测中,并结合2007-2011年上海市社会总抚养比数据建立新的预测模型,并用2012年数据对模型进行验证合格,可以用来预测未来几年上海市社会总抚养比,便于该市对未来经济的发展宏观调控.结果表明该预测方法是合理可行的,为其他相关预测提供了理论依据.     

缺失项二阶混料模型I最优设计

混料试验设计的应用热点已经从工农业转到医学领域,模型从通常的规范和中心多项式模型发展到适合应用的具有缺失项的模型,最优性准则中从参数估计准确出发的D最优研究较多,但是能准确预测响应面估计的I最优准则更适合研究混料设计,又因为需要复杂运算所以研究较少.现在利用MATLAB软件,在I最优准则下,对具有缺失项的二阶混料模型进行计算,区别于过去的线性优化,使用非线性优化,给出缺失项二阶混料模型的I最优设计重要的结论性定理.     

CPI数据预测计算的ANFIS模型结构探讨

依据CPI经济序列数据确定性混沌原理,探讨自适应神经模糊推理系统模型构造,并给出此类混沌数据列预测的ANFIS系统结构形式,进行CPI经济序列数据预测.并用实例拟合、预测数据证明:ANFIS模型是一种精度较高的混沌数据序列预报系统.为CPI数据预测提供了一种计算方法.     

基于多项式混沌拓展模型和bootstrap的结构可靠性分析方法

多项式混沌拓展(polynomial chaos expansion,PCE)模型现已发展为全局灵敏度分析的强大工具,却很少作为替代模型用于可靠性分析。针对该模型缺乏误差项从而很难构造主动学习函数来逐步更新的事实,在结构可靠性分析的框架下提出了基于PCE模型和bootstrap重抽样的仿真方法来计算失效概率。首先,对试验设计(experimental design)使用bootstrap重抽样步骤以刻画PCE模型的预测误差;其次,基于这个局部误差构造主动学习函数,通过不断填充试验设计以自适应地更新模型,直到能够精确地逼近真实的功能函数;最后,当PCE模型具有足够精确的拟合、预测能力,再使用蒙特卡洛仿真方法来计算失效概率。提出的平行加点策略既能在模型更新过程中找到改进模型拟合能力的"最好"的点,又考虑了模型拟合的计算量;而且,当失效概率的数量级较低时,PCE-bootstrap步骤与子集仿真(subset simulation)的结合能进一步加速失效概率估计量的收敛。本文方法将PCE模型在概率可靠性领域的应用从灵敏度分析延伸到了可靠性分析,同时,算例分析结果显示了该方法的精确性和高效性。     

基于马尔可夫模型的我国农业产值结构预测

利用马尔可夫模型对我国农业产值结构的预测问题进行了研究,并给出了其状态转移概率矩阵的估算模型与优化方法,应用此模型依据1999-2009年我国农业产值结构的统计数据,预测了农、林、牧、渔各业的产值结构.预测结果表明,该方法对农业产值结构的平均拟合误差为3.08%.用2010-2011年农业产值结构数据进行检验预测,结果表明模型可以有效地提高农业产值结构的预测精度.最后,用模型对2012-2015年我国农业产值结构进行了预测.     

多指标可加模型及在医疗费用预测中的应用

对医疗费用的建模分析与合理预测是医疗保险费用厘定的基础与根本.医疗费用中的高维附加信息在长期预测中具有重要作用.然而,传统的统计建模方法不适用于处理高维纵向数据下的医疗费用.本文提出部分线性多指标可加模型,对具有高维特征的纵向医疗费用数据进行拟合与预测,并且使用两种不同的降维估计方法进行模型估计,并将该模型应用于一组含...     

基于网络搜索数据和机器学习的票房预测模型

电影票房预测对于管理部门一直是一项重要而复杂的工作。电影票房相关变量复杂多变,且数据获取难度较大是制约当前研究的主要因素。相比之下,网络搜索数据是互联网公司发布的用于记录网民搜索行为的结构化数据,能客观及时反映事物的发展趋势。本研究建立了基于网络搜索数据的混合预测模型。首先,匹配与测试集最相似的训练数据构建最优训练集(OTS)。其次,应用帝国竞争算法(ICA)选择最小二乘支持向量机(LSSVM)的最佳参数组合。最后,使用优化模型进行预测。为了测试模型的效果,使用中国大陆上映的电影票房数据进行模拟实验。结果表明混合模型具有更高的预测精度。本研究所构建的模型适用于中国电影业的票房预测,可为有关部门提供决策参考。     

A novel method for non-probabilistic convex modelling based on data from practical engineering

In this paper, a novel method for non-probabilistic convex modelling with the bounds to precisely encircle all the data of uncertain parameters extracted from practical engineering is developed. The method is based on the traditional statistical method and the correlation analysis technique. Mean values and correlation coefficients of uncertain parameters are first calculated by utilizing the information of all the given data. Then, a simple yet effective optimization procedure is first introduced in the mathematical modelling process for uncertain parameters to obtain their precise bounds. This procedure works by optimizing the area of the convex model, at the same time, covering all the given data. Thus, the effective mathematical expression of the convex models are finally formulated. To test the prediction capability and generalization ability of the proposed convex modelling method, evaluation criteria, i.e. volume ratio, standard volume ratio, and prediction accuracy are established. The performance of the proposed method is systematically studied and compared with other existing competitive methods through test standards. The results demonstrate the effectiveness and efficiency of the present method.     

基于混合模型的利润驱动违约判别临界点研究

违约判别临界点是金融机构是否接受客户贷款申请的重要参考,合适的违约判别临界点对减少金融机构贷款损失实现稳健经营具有重要意义。本文研究的问题是如何保证计算客户违约概率的准确性,并找到利润最大化的违约判别临界点。本文的创新与特色:一是通过将多个不同类型的违约判别模型计算的客户违约概率进行加权平均,保证了计算客户违约概率的的整体准确性,避免了使用单一模型计算客户违约概率不准确的弊端;二是通过定义金融机构从贷款中获得利润的计算公式,以利润最大为目标,求解违约判别临界点,避免了现有计算临界点的方法如广义对称点估计和经验似然法等方法得到的临界点利润不是最大的弊端。研究发现:混合模型比单一模型的准确性高,AUC值显著提高;在人人贷数据集中本文的违约判别临界点下贷款利润远高于其他方法下临界点的利润。     

Adaptive network-based fuzzy inference system with leave-one-out cross-validation approach for prediction of surface roughness

The prediction of surface roughness is a challengeable problem. In order to improve the prediction accuracy in end milling process, an improved approach is proposed to model surface roughness with adaptive network-based fuzzy inference system (ANFIS) and leave-one-out cross-validation (LOO-CV) approach. This approach focuses on both architecture and parameter optimization. LOO-CV, which is an effective measure to evaluate the generalization capability of mode, is employed to find the most suitable membership function and the optimal rule base of ANFIS model for the issue of surface roughness prediction. To find the optimal rule base of ANFIS, a new “top down” rules reduction method is suggested. Three machining parameters, the spindle speed, feed rate and depth of cut are used as inputs in the model. Based on the same experimental data, the predictive results of ANFIS with LOO-CV are compared with the results reported recently in the literature and ANFIS with clustering methods. The comparisons indicate that the presented approach outperforms the opponent methods, and the prediction accuracy can be improved to 96.38%. ANFIS with LOO-CV approach is an effective approach for prediction of surface roughness in end milling process.     

A gray model for increasing sequences with nonhomogeneous index trends based on fractional‐order accumulation

By comparing the class ratio deviation and restoring error of first‐order accumulation with that of fractional‐order accumulation, a gray model for monotonically increasing sequences can obtain optimal simulation accuracy via selecting a proper cumulative order. In this study, a gray model for increasing sequences with nonhomogeneous index trends based on fractional‐order accumulation is proposed. To reduce the modeling error caused by the background value and to improve the prediction accuracy of the model, an optimized model using the 3/8 Simpson formula is constructed. Finally, the 2 proposed models are used to predict the total energy consumption in China and the monthly sales of new products in an enterprise. Compared with the GM(1,1) model based on fractional‐order accumulation, the proposed model exhibits better simulation and prediction accuracy.     

A bioeconomic differential algebraic predator–prey model with nonlinear prey harvesting

In this paper, we propose a bioeconomic differential algebraic predator–prey model with Holling type II functional response and nonlinear prey harvesting. As the nonlinear prey harvesting is introduced, the proposed model displays a complex dynamics in the predator–prey plane. Taking into account of the economic factor, our predator–prey system is established by bioeconomic differential algebraic equations. The effect of economic profit on the proposed model is analyzed by viewing it as a bifurcation parameter. By jointly using the normal form of differential algebraic models and the bifurcation theory, the stability and bifurcations (singularity induced bifurcation, Hopf bifurcation) are discussed. These results obtained here reveal richer dynamics of the bioeconomic differential algebraic predator–prey model with nonlinear prey harvesting, and suggest a guidance for harvesting in the practical word. Finally, numerical simulations are given to demonstrate the results.     

The development of hybrid ES-FE-SEA method for mid-frequency vibration analysis of complex built-up structure

The beam-plate structures and its complex composite structures are widely used in practical engineering. Recently, a hybrid FE-SEA method was applied for the vibration analysis of beam-plate structures in mid-frequency regime. However, the accuracy of prediction is still low, due to “overly-stiff” feature of embedded conventional FEM. In this work, a hybrid ES-FE-SEA model of a beam-plate built-up structure is developed for response prediction in vibration testing, in which, the edge-based smoothed technique is applied in the 3D beam structure FEM model to soften the whole system stiffness. Then, combined with SEA, this ES-FEM is embedded to achieve a hybrid ES-FE-SEA framework, to improve the accuracy of mid-frequency response predictions of the complex built-up structure. In hybrid ES-FE-SEA model, the plate structure which has a higher model density is considered as a statistical subsystem and modeled statistically using statistical energy analysis (SEA). The beam structure with a relative lower model density is modeled deterministically using edge-based smooth finite element method (ES-FEM). The coupling between these two different types of subsystems is achieved through the diffuse field reciprocity relation. The acceleration loads are applied to the model. The results obtained by the ES-FE-SEA and FE-SEA are compared. It is found that the hybrid ES-FE-SEA method is reliable for the mid-frequency vibration problems of the beam-plate built-up structure. The proposed ES-FE-SEA is verified by various numerical examples.     

基于面板logit模型的上市公司财务困境预测

目前关于财务困境预测的研究大多是局限于截面数据的静态计量和统计模型,忽视了公司的财务状况是不断变化的事实.为了揭示公司财务状况的变化过程,利用面板数据建立了panel logit概率模型.研究结果表明,panel logit模型在预测准确度方面优于普通的logit模型.     

基于小波变换和改进萤火虫算法优化极限学习机的短期负荷预测

提出了一种基于小波变换和改进萤火虫优化极限学习机的短期负荷预测方法.通过小波分解和重构,对原始负荷序列进行降噪;在模型训练阶段利用改进的萤火虫算法优化极限学习机参数,获得各序列的最优模型;针对各子序列分别预测叠加得到最终预测值.通过在两种时间尺度的数据序列上进行数值计算,与传统的ARMA、BP神经网络、支持向量机及LSSVM等多种经典预测模型相比,模型预测效果更优.     

基于卷积神经网络气动力降阶模型的翼型优化方法

针对非线性大扰动翼型气动力优化问题,提出了基于卷积神经网络气动力降阶模型的优化方法.该方法用不同形状参数下翼型的气动力数据作为训练信号,训练卷积神经网络翼型气动力降阶模型.采用该气动力降阶模型,以最大升阻比为目标,对翼型进行优化,结果表明该方法可用于大扰动下翼型气动力的预测和优化.该文同时还讨论了池化法和径向基法的训练...