多目标预报系统的Fuzzy 优选识别模型及其应用

对多目标预报系统的Ｆｕｚｙ优选识别模型进行了分析和研究,并建立了山东省曲阜市第二代棉铃虫发生量的Ｆｕｚｙ优选识别模型。对历史资料进行回代验证,其历史拟合率高达９１．６７％。将１９９４年的观测数据作为独立样本进行试报,预测结果与实际一致。     

结合神经元网络和模糊专家系统进行电力短期负荷预测

结合人工神经元网络（ＡＮＮ）和模糊专家系统进行负荷预测．给出了径向基函数（ＲＢＦ）网络的结构,并采用正交最小平方法（ＯＬＳ）选取ＲＢＦ中心．先用ＡＮＮ进行基本负荷预测,然后考虑天气变化和假日因素所引起的负荷变化,利用模糊专家系统进行负荷调整．文中还把日期划分为５类．测试结果表明,该方法具有较高的精度和较快的速度．     

基于时空特性的短时交通流预测模型

根据交通流的历史周期性和空间相关性,文中综合SARIMA模型在历史周期性预测上的优势和RBF模型在空间相关性预测上的优势,提出了SARIMA-RBF模型.该模型采用SARIMA模型通过历史数据预测下一时刻的交通流,然后将预测值与该点上下游关联的交通流数据相结合,采用RBF神经网络模型得出输出值,并将该输出值作为SARIMARBF模型对下一时刻交通流的预测结果.实验结果表明,该模型因同时考虑了交通流的历史周期性和空间相关性,相比SARIMA模型和RBF模型具有更好的交通流预测效果.     

Exploiting cooperative multipath –Doppler diversity in relay-assisted high-speed train communications

Severe doubly selective channel fading, or the time- and frequency-selective fading, prevailing in high- speed train （HST） communications, has been a critical issue that hinders the improvement of system efficiency and reli- ability. In this paper, a relay-assisted HST communication system is considered, and a low-complexity but effective scheme is proposed to cooperatively exploit the joint mul- tipath-Doppler diversity of both the direct link and the relay link and thus to mitigate the negative effect of channel fading and improve the system reliability. In particular, using a special precoding structure followed by two-dimensional Fourier transform, the transmit signal is generated which in effect parallelizes the doubly selective fading channel. As a result, in the transform domain, pointwise combining and equalization, which are of linear complexity, become feasi- ble and diversity gain is obtained. The maximal diversity order is then analyzed under the amplify-and-forward relay settings, which reveals that by cooperative relaying the joint multipath-Doppler diversity of both links can be well exploited. Simulation results verified our analysis.     

连续参数小波网络在机械状态预报中的应用

研究了连续参数小波网络的原理及其构造,通过仿真数据的比较分析,说明了连续小波网络较传统ＢＰ网络有更快的收敛速度和更好的非线性逼近能力。同时,在用连续参数小波网络对实际机组振动峰峰值数据的预报中,收到了满意的效果。     

采样间隔非均匀情况下的季节模型及其贝叶斯预测

主要给出了采样间隔非均匀情况下的季节因子动态线性模型、季节效应动态线性模型和Ｆｏｕｒｉｅｒ形式的季节效应模型,当观测误差方差Ｖｔｉ已知时,给出了它们的先验分布、预测分布和后验分布的修正递推算法。     

直接建模与累加建模灰色模型特性的比较

研究了直接建模与累加建模所建立的灰色模型的模型特性,得到了一些有价值的结论－这些结论有助于灰色预测模型在有关领域中的应用－     

基于时间序列的市场需求预测模型研究

在研究时间序列预测模型的基础上,提出了组合预测模型,并对模型参数进行了优化处理.用于预算未来某时期内市场供求趋势,从而为企业的决策提供依据.     

An aggregate sales model for consumer durables incorporating a time‐varying mean replacement age

Forecasting category or industry sales is a vital component of a company's planning and control activities. Sales for most mature durable product categories are dominated by replacement purchases. Previous sales models which explicitly incorporate a component of sales due to replacement assume there is an age distribution for replacements of existing units which remains constant over time. However, there is evidence that changes in factors such as product reliability/durability, price, repair costs, scrapping values, styling and economic conditions will result in changes in the mean replacement age of units. This paper develops a model for such time‐varying replacement behaviour and empirically tests it in the Australian automotive industry. Both longitudinal census data and the empirical analysis of the replacement sales model confirm that there has been a substantial increase in the average aggregate replacement age for motor vehicles over the past 20 years. Further, much of this variation could be explained by real price increases and a linear temporal trend. Consequently, the time‐varying model significantly outperformed previous models both in terms of fitting and forecasting the sales data. Copyright © 2001 John Wiley & Sons, Ltd.     

Alternative regime switching models for forecasting inflation

US inflation appears to undergo shifts in its mean level and variability. We evaluate the performance of three useful models for capturing such shifts. The models studied are the Markov switching models, state space models with heavy‐tailed errors, and state space models with compound error distributions. Our study shows that all three models have very similar performance when evaluated in terms of the mean squared or mean absolute forecast errors. However, the latter two models are considerably more parsimonious, and easily beat the more profligately parameterized Markov switching models in terms of model selection criteria, such as the AIC or the SBC. Thus, these may serve as useful continuous alternatives to the popular discrete Markov switching models for capturing shifts in time series. Copyright © 2001 John Wiley & Sons, Ltd.